Signal Transduction in the Retina

SIGNAL TRANSDUCTION IN THE RETINA 7315.indb 1 11/15/07 1:36:45 PM M ETHODS IN SIGNAL TRANSDUCTION METHODS IN SIG...

Author: Steven J. Fliesler | Oleg G. Kisselev

376 downloads 1352 Views 10MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

SIGNAL TRANSDUCTION IN THE RETINA

7315.indb 1

11/15/07 1:36:45 PM

M

ETHODS

IN SIGNAL TRANSDUCTION

METHODS IN SIGNAL TRANSDUCTION SERIES Joseph Eichberg, Jr., Series Editor

Published Titles Lipid Second Messengers, Suzanne G. Laychock and Ronald P. Rubin G Proteins: Techniques of Analysis, David R. Manning Signaling Through Cell Adhesion Molecules, Jun-Lin Guan G Protein-Coupled Receptors, Tatsuya Haga and Gabriel Berstein Calcium Signaling, James W. Putney, Jr. G Protein-Coupled Receptors: Structure, Function, and Ligand Screening, Tatsuya Haga and Shigeki Takeda Calcium Signaling, Second Edition James W. Putney, Jr. Analysis of Growth Factor Signaling in Embryos Malcolm Whitman and Amy K. Sater

7315.indb 2

11/15/07 1:36:46 PM

SIGNAL TRANSDUCTION IN THE RETINA Edited by

Steven J. Fliesler & Oleg G. Kisselev

Boca Raton London New York

CRC Press is an imprint of the Taylor & Francis Group, an informa business

7315.indb 3

11/15/07 1:36:46 PM

CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 2008 by Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group, an Informa business No claim to original U.S. Government works Printed in the United States of America on acid-free paper 10 9 8 7 6 5 4 3 2 1 International Standard Book Number-13: 978-0-8493-7315-2 (Hardcover) This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the consequences of their use. No part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www. copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC) 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data Signal transduction in the retina / [edited by] Steven J. Fliesler and Oleg Kisselev. p. ; cm. -- (Methods in signal transduction series) “A CRC title.” Includes bibliographical references and index. ISBN 978-0-8493-7315-2 (alk. paper) 1. Retina--Physiology. 2. Cellular signal transduction. I. Fliesler, Steven J. II. Kisselev, Oleg. III. Title. IV. Series: Methods in signal transduction. [DNLM: 1. Phototransduction--physiology. 2. Retina--physiology. 3. Rods (Retina)--physiology. 4. Vertebrates. WW 270 S5784 2008] QP479.S59 2008 612.8’43--dc22

2007027750

Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com and the CRC Press Web site at http://www.crcpress.com

T&F_LOC_A_Master.indd 1 7315.indb 4

7/11/07 7:58:14 11/15/07 1:36:46 AM PM

The Editors dedicate this volume to Dr. Paul A. Hargrave, in honor of his numerous and significant contributions to the field of vision science, particularly in regard to the biochemistry and structure of rhodopsin and the current understanding of the phototransduction cascade in retinal rod cells.

7315.indb 5

11/15/07 1:36:47 PM

7315.indb 6

11/15/07 1:36:47 PM

Contents Series Preface.............................................................................................................xi Preface.................................................................................................................... xiii Editors....................................................................................................................... xv List of Contributors.................................................................................................xvii Section 1

Vertebrate Visual Phototransduction

Chapter 1 Biophysical Approaches to the Study of G-Protein Structure and Function..........................................................................................3 Anita M. Preininger, William M. Oldham, and Heidi E. Hamm Chapter 2 Photoactivation of Rhodopsin and Signal Transfer to Transducin........................................................................................... 33 Oleg G. Kisselev Chapter 3 How Rod Arrestin Achieved Perfection: Regulation of Its Availability and Binding Selectivity................................................... 55 Vsevolod V. Gurevich, Susan M. Hanson, Eugenia V. Gurevich, and Sergey A. Vishnivetskiy Chapter 4 Function and Regulation of PDE6....................................................... 89 Michael Natochin, Hakim Muradov, and Nikolai O. Artemyev Chapter 5 Biochemical Characterization of Phototransduction RGS9-1–GAP Complex.......................................................................99 Qiong Wang and Theodore G. Wensel Chapter 6 Guanylate Cyclase-Based Signaling in Photoreceptors and Retina.......................................................................................... 121 Karl-Wilhelm Koch and Andreas Helten Chapter 7 Transgenic Strategies for Analysis of Photoreceptor Function......... 145 Janis Lem and Kibibi Rwayitare

vii

7315.indb 7

11/15/07 1:36:47 PM

viii

Section 2

Signal Transduction in the Retina

Vertebrate Nonvisual Phototransduction

Chapter 8 Melanopsin Signaling and Nonvisual Ocular Photoreception.......... 165 Sowmya V. Yelamanchili, Victoria Piamonte, Surendra Kumar Nayak, Nobushige Tanaka, Quansheng Zhu, Kacee Jones, Hiep Le, and Satchidananda Panda Section 3

Invertebrate Visual Phototransduction

Chapter 9 Phototransduction in Drosophila: Use of Microarrays in Cloning Genes Identified by Chemically Induced Mutations Causing ERG Defects........................................................................ 195 Hung-Tat Leung, Lingling An, Julie Tseng-Crank, Eunju Kim, Eric L. Harness, Ying Zhou, Junko Kitamoto, Guohua Li, Rebecca W. Doerge, and William L. Pak Section 4

Insulin Receptor-Based Signaling in the Vertebrate Retina

Chapter 10 Insulin Receptor-Based Signaling in the Retina............................... 221 Patrice E. Fort, Ravi S.J. Singh,Mandy K. Losiewicz, and Thomas W. Gardner Chapter 11 Probing the Interactions between the Retinal Insulin Receptor and Its Downstream Effectors........................................................... 237 Raju V.S. Rajala and Robert E. Anderson Section 5

Signal Transduction in Vertebrate Retinal Development and Vascular Homeostasis

Chapter 12 Signal Transduction in Retinal Progenitor Cells............................... 269 Branden R. Nelson, Susan Hayes, Byron H. Hartman, and Thomas A. Reh Chapter 13 Probing the Hedgehog Pathway in Retinal Development................. 299 Brian McNeill, Dana Wall, Shawn Beug, and Valerie A. Wallace Chapter 14 Thrombospondin-1 Signal Transduction and Retinal Vascular Homeostasis...................................................................................... 329 Nader Sheibani, Yixin Tang, Terri DiMaio,Shuji Kondo, Elizabeth A. Scheef, and Christine M. Sorenson

7315.indb 8

11/15/07 1:36:47 PM

ix

Contents

Section 6

Lipid Mediators and Signaling in the RPE

Chapter 15 Lipidomic Approaches to Neuroprotection Signaling in the Retinal Pigment Epithelium.................................................... 349 Nicolas G. Bazan, Victor L. Marcheselli, Yan Lu, Song Hong, and Fannie Jackson Index....................................................................................................................... 375

7315.indb 9

11/15/07 1:36:48 PM

7315.indb 10

11/15/07 1:36:48 PM

Series Preface The concept of signal transduction at the cellular level is now established as a cornerstone of the biological sciences. Cells sense and react to environmental cues by means of a vast panoply of signaling pathways and cascades. While the steady accretion of knowledge regarding signal transduction mechanisms is continuing to add layers of complexity, this greater depth of understanding has also provided remarkable insights into how healthy cells respond to extracellular and intracellular stimuli and how these responses can malfunction in many disease states. Central to advances in unraveling signal transduction is the development of new methods and refinement of existing ones. Progress in the field relies upon an integrated approach that utilizes techniques drawn from cell and molecular biology, biochemistry, genetics, immunology, and computational biology. The overall aim of this series is to collate and continually update the wealth of methodology now available for research into many aspects of signal transduction. Each volume is assembled by one or more editors who are leaders in their specialty. Their guiding principle is to recruit knowledgeable authors who will present procedures and protocols in a critical yet reader-friendly format. Our goal is to assure that each volume will be of maximum practical value to a broad audience, including students, seasoned investigators, and researchers who are new to the field. The retina has long been a favorite system for the study of signal transduction mechanisms because of its accessibility and the relative abundance of several of the components of the visual phototransduction pathway. Its investigation has yielded valuable insights into the molecular transformation that accompany the transmission of light signals. The editors of this volume have brought together a distinguished group of authors who build on existing knowledge to describe the latest methodological innovations for the exploration of phototransduction at the level of individual molecules. While the chief focus is on vertebrate photo-transduction, aspects of nonvisual and invertebrate phototransduction are also covered. Additional chapters deal with techniques used to investigate the rapidly developing area of insulin-mediated signaling in the retina, and with approaches to elucidate the role of signal transduction in the development of retina and supporting structures. A particularly attractive feature is the inclusion of protocols that provide detailed guidance in applying a variety of experimental methods to study retinal signal transduction. Without question, this volume will be a valuable reference to all investigators who are active or interested in this field. Joseph Eichberg, Ph.D. Advisory Editor for the Series

xi

7315.indb 11

11/15/07 1:36:48 PM

7315.indb 12

11/15/07 1:36:48 PM

Preface In the current “postgenomic era,” there is increasing recognition of the need for integrated approaches to study and understand complex biological systems and signaling networks. The retina—an anatomically and functionally unique part of the central nervous system, responsible for the detection and initial processing of visual information—is illustrative of this. An integrated knowledge of the biochemistry, cell biology, physiology, and physics of phototransduction, as well as postphotoreceptor visual transduction processes, has evolved over the past century, with the finer details becoming apparent particularly within the past decade. The retina is an extremely useful biological system amenable to experimental manipulation in vivo as well as in vitro, affording an accessible model with which to understand individual cellular signaling systems down to the level of molecular interactions at atomic resolution, as well as more complex issues of pathway regulation and the integration of signaling networks that impact cellular and tissue responses, ultimately resulting in visual perception. The present volume, comprised of fifteen chapters in six sections, brings together a number of internationally recognized authorities in disciplines pertinent to the study of signal transduction in the retina. Each chapter presents a brief overview of the background and current state of knowledge in a particular area relevant to the broader topic of retinal signal transduction, along with detailed information regarding specific methodology for obtaining the primary data necessary to understand the molecular and cellular processes being examined. Because more is known about the rhodopsin-based phototransduction pathway in vertebrate retinal rod cells than in almost any other biological system, and this dominates signaling processes in the retina, a substantial portion of this volume is devoted to that topic. In addition, a diversity of other signaling mechanisms and systems are covered, affording the reader a resource for evaluating the similarities and differences between these systems and the specific research strategies employed for studying them. Section 1 deals with the molecular mechanisms of vertebrate phototransduction, dissecting the major components of the phototransduction cascade in rod cells and proteins involved in its regulation. The chapters in this section emphasize the breadth of knowledge accumulated in the past decade, especially with regard to determination of the molecular structure of phototransduction cascade components at atomic resolution, as well as the use of transgenic strategies. State-of-the-art approaches for the study of molecular interactions in multiprotein complexes, as well as novel cellbased strategies aimed at understanding the mechanisms of signal shut-off and light adaptation, are presented. Section 2 focuses on the more recently emerging field of nonvisual phototransduction. Methods for assessing the roles of melanopsin in regulation of the circadian clock and in adaptive photoresponses are described. Section 3 provides a chapter devoted to essential methods for studying phototransduction in the invertebrate retina, using Drosophila as the biological system xiii

7315.indb 13

11/15/07 1:36:48 PM

xiv


of choice. Thus, the reader will be able to compare and contrast the juxtaposing processes of visual signaling in vertebrates versus invertebrates. Section 4 focuses on experimental studies of insulin-based signaling in the retina, both in the outer retina (photoreceptors, per se) as well as inner retinal cells. In addition, insulin receptor structure and ligand-binding specificity as well as mechanisms of downstream signaling are described. Section 5 presents current methodological approaches relevant to retinal development, including cellular signaling in retinal progenitor cells, and cell–cell communications in developing retina. Because neovascularization is considered an increasingly important factor in various human degenerative retina diseases, particularly those that accompany diabetes and aging, this section also addresses experimental approaches for studying vascular homeostasis. Section 6 deals with recent developments in the field of lipid-derived mediators, particularly neuroprotectins and the participation of the retinal pigment epithelium in neuronal survival in the retina. Now in the twenty-first century, we are just beginning to understand the enormous diversity and complexity of signaling processes in the retina. The methodologies and experimental approaches described in this volume have already yielded key fundamental information regarding the cellular and molecular mechanisms that underlie normal retinal physiology. In addition, they have the potential to provide new clues toward elucidating the mechanisms involved in retinal disease processes as well as the development of novel therapeutic approaches for preventing, arresting, and modulating those disease processes and promoting cellular survival and retention of function. Steven J. Fliesler Oleg G. Kisselev

7315.indb 14

11/15/07 1:36:48 PM

Editors Steven J. Fliesler is a professor and director of research, Department of Ophthalmology (Saint Louis University Eye Institute), as well as a professor in the Department of Pharmacological and Physiological Science at Saint Louis University School of Medicine, in St. Louis, Missouri. He earned a B.A. degree in biochemistry, with a minor in chemistry, from the University of California–Berkeley in 1973, and a Ph.D. degree in biochemistry from Rice University (Houston, Texas) in 1980. Following a three-year postdoctoral research fellowship on retinal lipid metabolism at the Cullen Eye Institute, Baylor College of Medicine (Houston, Texas), Fliesler pursued studies funded by the National Institutes of Health (NIH) on glycoprotein metabolism and photoreceptor membrane assembly in the retina as a research assistant professor at the Cullen Eye Institute. In 1985, Fliesler moved to the Bascom Palmer Eye Institute, with a joint appointment as an assistant professor in the Department of Biochemistry and Molecular Biology and in the Program in Neuroscience at the University of Miami School of Medicine, where he continued his studies on glycoprotein metabolism in the retina. In 1988, he was appointed associate professor at the Bethesda Eye Institute (now Saint Louis University Eye Institute), with a secondary appointment as associate professor in the E.A. Doisy Department of Biochemistry and Molecular Biology and the Cell and Molecular Biology Graduate Program at Saint Louis University School of Medicine (St. Louis, Missouri). Subsequently, Fliesler was promoted (in 1994) to professor in the Department of Ophthalmology, with a secondary appointment (in 2000) as professor in the Department of Pharmacological and Physiological Science at Saint Louis University School of Medicine. Fliesler’s research program encompasses studies on the relationship between cellular metabolism and the establishment and preservation of retinal structure and function, especially in regard to retinal rod photoreceptor cells. A major focus of his research has been cholesterol metabolism and dyslipidemias caused by inborn errors in cholesterol metabolism. In addition, he has a long-standing interest in animal models of human hereditary retinal degenerations, particular those involving defective membrane transport and assembly in retinal photoreceptor cells. Fliesler is the author or coauthor of over 90 peer-reviewed publications, book chapters, and review articles dealing largely with retinal cell biology and glycoprotein and lipid metabolism, and has delivered more than 150 presentations at national and international scientific meetings, colleges and universities, and specialty scientific and biomedical symposia. In addition to editing this volume, Fliesler is the editor of Sterols and Oxysterols: Chemistry, Biology and Pathobiology, a multiauthor volume in the “Recent Research Developments in Biochemistry” series, published by Research Signpost (Kerala, India) in 2002. Fliesler is an editorial board member, Retina and Choroid Section editor, and “Focus on Molecules” feature editor for Experimental Eye Research, and also serves on the editorial board of the Journal of Lipid Research. In addition, he also serves regularly as a reviewer for a variety of federal, state, and private grant-funding agencies, including the National Eye Institute, the Foundation xv

7315.indb 15

11/15/07 1:36:49 PM

xvi


Fighting Blindness, and Fight for Sight. Fliesler is the recipient of numerous research grant awards, both from federal and private funding agencies, including the National Institutes of Health, the Foundation Fighting Blindness, the March of Dimes Foundation, and Research to Prevent Blindness. In 2007, he was selected as the recipient of a Senior Scientific Investigator Award from Research to Prevent Blindness. Oleg G. Kisselev is currently an associate professor in the Department of Ophthalmology (Saint Louis University Eye Institute), with a secondary faculty appointment in the Edward A. Doisy Department of Biochemistry and Molecular Biology at Saint Louis University School of Medicine. Kisselev obtained his undergraduate and Ph.D. degrees in biochemistry at Lomonosov Moscow State University, Moscow. In 1992, he emigrated to the United States to pursue postdoctoral training in the laboratory of N. Gautam in the Department of Anesthesiology at Washington University School of Medicine in St. Louis. He later became an American Heart Association Fellow and, in 1997, he received his first faculty appointment as an instructor in that department with a secondary appointment in the Center for Molecular Design headed by Professor Garland R. Marshall at Washington University. In 1999, Kisselev established his own independent research program upon accepting a faculty position in the Department of Ophthalmology (Saint Louis University Eye Institute) of Saint Louis University School of Medicine. He was promoted to his current rank of associate professor in 2007. Kisselev’s research interests are in the biology of sensory signaling utilizing universal mechanisms of transmembrane signal transduction mediated by heterotrimeric GTP-binding proteins, and G-protein-coupled cell surface receptors. He has made seminal contributions to the current state of knowledge regarding the role of individual G-protein subunits in determining the specificity of G-protein signaling and the mechanism of receptor-catalyzed G-protein activation, especially in the vertebrate visual system. His studies of interactions between phototransduction proteins using high-resolution nuclear magnetic resonance (NMR) methods have helped to elucidate the dynamics of the phototransduction machinery at atomic resolution, and have provided essential refinements to the mechanism of visual signal transduction. Kisselev is an author of more than 25 scientific publications including reviews and book chapters dealing with the biochemistry and structural biology of G-proteins. He has received funding for his research program from the National Institutes of Health and the American Heart Association. In addition, he received a William and Mary Greve Scholar Award in 2002 from Research to Prevent Blindness.

7315.indb 16

11/15/07 1:36:49 PM

List of Contributors Lingling An Department of Statistics Purdue University West Lafayette, Indiana Robert E. Anderson Departments of Ophthalmology and Cell Biology, and Dean A. McGee Eye Institute Health Sciences Center University of Oklahoma Oklahoma City, Oklahoma Nikolai O. Artemyev Department of Molecular Physiology and Biophysics College of Medicine University of Iowa Iowa City, Iowa Nicolas G. Bazan Neuroscience Center of Excellence and Department of Ophthalmology Health Sciences Center School of Medicine Louisiana State University New Orleans, Louisiana Shawn Beug Molecular Medicine Program and Vision Program Ottawa Health Research Institute and Department of Biochemistry, Microbiology and Immunology University of Ottawa Ottawa, Ontario, Canada

Terri DiMaio Department of Ophthalmology and Visual Sciences School of Medicine and Public Health University of Wisconsin Madison, Wisconsin Rebecca W. Doerge Department of Statistics Purdue University West Lafayette, Indiana Patrice E. Fort Departments of Ophthalmology and Cellular and Molecular Physiology College of Medicine Pennsylvania State University Hershey, Pennsylvania Thomas W. Gardner Departments of Ophthalmology and Cellular and Molecular Physiology College of Medicine Pennsylvania State University Hershey, Pennsylvania Eugenia V. Gurevich Department of Pharmacology School of Medicine Vanderbilt University Nashville, Tennessee Vsevolod V. Gurevich Department of Pharmacology School of Medicine Vanderbilt University Nashville, Tennessee

xvii

7315.indb 17

11/15/07 1:36:49 PM

xviii

Heidi E. Hamm Department of Pharmacology School of Medicine Vanderbilt University Nashville, Tennessee Susan M. Hanson Department of Pharmacology School of Medicine Vanderbilt University Nashville, Tennessee Eric L. Harness Department of Biological Sciences Purdue University West Lafayette, Indiana Byron Hartman Department of Biological Structure School of Medicine University of Washington Seattle, Washington Susan Hayes Department of Biological Structure School of Medicine University of Washington Seattle, Washington Andreas Helten Biochemistry Group Department of Biology and Environmental Sciences Carl von Ossietzky University Oldenburg, Germany Song Hong Neuroscience Center of Excellence and Department of Ophthalmology School of Medicine Health Sciences Center Louisiana State University New Orleans, Louisiana

7315.indb 18


Fannie Jackson Neuroscience Center of Excellence and Department of Ophthalmology School of Medicine Health Sciences Center Louisiana State University New Orleans, Louisiana Kacee Jones Regulatory Biology Laboratory The Salk Institute of Biological Sciences La Jolla, California Eunju Kim Department of Biological Sciences Purdue University West Lafayette, Indiana Oleg G. Kisselev Departments of Ophthalmology and Biochemistry and Molecular Biology School of Medicine Saint Louis University Saint Louis, Missouri Junko Kitamoto Department of Biological Sciences Purdue University West Lafayette, Indiana Karl-Wilhelm Koch Biochemistry Group Department of Biology and Environmental Sciences Carl von Ossietzky University Oldenburg, Germany Shuji Kondo Department of Pediatrics School of Medicine and Public Health University of Wisconsin Madison, Wisconsin

11/15/07 1:36:49 PM

xix

List of Contributors

Hiep Le Regulatory Biology Laboratory The Salk Institute of Biological Sciences La Jolla, California Janis Lem Molecular Cardiology Research Institute and Department of Ophthalmology Tufts-New England Medical Center Tufts University Boston, Massachusetts Sackler School of Graduate Biomedical Sciences School of Medicine Hung-Tat Leung Department of Biological Sciences Purdue University West Lafayette, Indiana Guohua Li Department of Biological Sciences Purdue University West Lafayette, Indiana Mandy K. Losiewicz Departments of Ophthalmology and Cellular and Molecular Physiology College of Medicine Pennsylvania State University Hershey, Pennsylvania Yan Lu Neuroscience Center of Excellence and Department of Ophthalmology School of Medicine Louisiana State University New Orleans, Louisiana Victor L. Marcheselli Neuroscience Center of Excellence and Department of Ophthalmology

7315.indb 19

School of Medicine Louisiana State University New Orleans, Louisiana Brian McNeill Ottawa Health Research Institute and Department of Biochemistry, Microbiology, and Immunology University of Ottawa Ottawa, Ontario Canada Hakim Muradov Department of Molecular Physiology and Biophysics College of Medicine University of Iowa Iowa City, Iowa Michael Natochin Department of Molecular Physiology and Biophysics College of Medicine University of Iowa Iowa City, Iowa Surendra Kuman Nayak Genomics Institute of Novartis Research Foundation San Diego, California Branden Nelson Department of Biological Structure School of Medicine University of Washington Seattle, Washington William M. Oldham Department of Pharmacology School of Medicine Vanderbilt University Nashville, Tennessee

11/15/07 1:36:49 PM

xx

William L. Pak Department of Biological Sciences School of Medicine Purdue University West Lafayette, Indiana Satchidananda Panda Regulatory Biology Laboratory The Salk Institute of Biological Sciences La Jolla, California Victoria Piamonte Genomics Institute of Novartis Research Foundation San Diego, California Anita M. Preininger Department of Pharmacology School of Medicine Vanderbilt University Nashville, Tennessee Raju V. S. Rajala Departments of Ophthalmology and Cell Biology, and Dean A. McGee Eye Institute Health Sciences Center University of Oklahoma Oklahoma City, Oklahoma Thomas Reh Department of Biological Structure School of Medicine University of Washington Seattle, Washington Kibibi Rwayitare Molecular Cardiology Research Institute and Department of Ophthalmology Tufts-New England Medical Center School of Medicine Tufts University Boston, Massachusetts

7315.indb 20


Elizabeth A. Scheef Department of Ophthalmology and Visual Sciences School of Medicine and Public Health University of Wisconsin Madison, Wisconsin Nader Sheibani Departments of Ophthalmology and Visual Sciences and Pharmacology School of Medicine and Public Health University of Wisconsin Madison, Wisconsin Ravi S. J. Singh Departments of Ophthalmology and Cellular and Molecular Physiology College of Medicine Pennsylvania State University Hershey, Pennsylvania Christine M. Sorenson Department of Pediatrics School of Medicine and Public Health University of Wisconsin Madison, Wisconsin Nobushige Tanaka Regulatory Biology Laboratory The Salk Institute of Biological Sciences La Jolla, California Yixin Tang Department of Ophthalmology and Visual Sciences School of Medicine and Public Health University of Wisconsin Madison, Wisconsin Julie Tseng-Crank Department of Biological Sciences Purdue University West Lafayette, Indiana

11/15/07 1:36:50 PM

xxi

List of Contributors

Sergey A. Vishnivetskiy Department of Pharmacology School of Medicine Vanderbilt University Nashville, Tennessee Dana Wall Ottawa Health Research Institute and Department of Biochemistry, Microbiology, and Immunology University of Ottawa Ottawa, Ontario, Canada Valerie A. Wallace Ottawa Health Research Institute and Department of Biochemistry, Microbiology, and Immunology University of Ottawa Ottawa, Ontario, Canada Qiong Wang Department of Biochemistry and Molecular Biology College of Medicine Baylor University Houston, Texas

7315.indb 21

Theodore G. Wensel Department of Biochemistry and Molecular Biology College of Medicine Baylor University Houston, Texas Sowmya V. Yelamanchili Regulatory Biology Laboratory The Salk Institute of Biological Sciences La Jolla, California Ying Zhou Department of Biological Sciences Purdue University West Lafayette, Indiana Quansheng Zhu Regulatory Biology Laboratory The Salk Institute of Biological Sciences La Jolla, California

11/15/07 1:36:50 PM

7315.indb 22

11/15/07 1:36:50 PM

1 Vertebrate Visual Phototransduction

7315.indb 1

11/15/07 1:37:16 PM

7315.indb 2

11/15/07 1:37:17 PM

1

Biophysical Approaches to the Study of G-Protein Structure and Function Anita M. Preininger, William M. Oldham, and Heidi E. Hamm

Contents 1.1

Introduction......................................................................................................4 1.1.1 Overall Strategy....................................................................................5 1.2 Protein Expression and Purification.................................................................6 1.2.1 Expression and Purification of Gαi Hexa I Proteins.............................7 1.2.2 Rhodopsin and Gβγ Preparation......................................................... 10 1.3 Functional Assays.......................................................................................... 11 1.3.1 Intrinsic Tryptophan Fluorescence..................................................... 11 1.3.2 Basal Nucleotide Exchange................................................................. 12 1.3.3 Receptor-Mediated GTPγS Binding................................................... 15 1.3.4 Receptor Binding................................................................................ 17 1.4 Fluorescent Probes for Analysis of Conformational Changes in GPCR Signaling........................................................................................................ 18 1.4.1 Fluorescent Labeling of Gα Subunits................................................. 18 1.5 Fluorescence Measurement of Protein Dynamics......................................... 21 1.5.1 Analysis of Conformational Changes in Gα Subunits Using Fluorescent Techniques....................................................................... 21 1.5.2 Subunit Interactions as Measured by Fluorescence Resonance Energy Transfer (FRET)..................................................................... 22 1.5.3 Fluorescence Polarization Measurement of Protein Dynamics..........24 1.6 Analysis of Conformational Changes Using EPR......................................... 27 1.6.1 Spin-Labeling of Gαi Hexa I Proteins................................................28 1.6.2 EPR Analysis of Conformational Changes in Gαi Hexa I Proteins............................................................................................... 29 References................................................................................................................. 31

3

7315.indb 3

11/19/07 8:25:36 PM

4


1.1 Introduction Crystallographic studies have revealed much of what we know about G-protein structure, thus allowing us to better understand how these proteins function. The structural studies of the Gαi and Gαt-GDP, GTPγS, GDP-AlF4−, and Gβγ bound forms (1–6) have revealed many details about mechanisms of G-protein activation. The crystal structures of heterotrimeric Gi and Gt (4,6) and activated subunits have allowed the identification of subunit binding sites and have shed light on how GTP binding leads to subunit dissociation and effector activation. Comparison of the crystal structures in the active and inactive states reveal regions of conformational flexibility within the switch regions that undergo specific changes upon activation and deactivation. When Gα is activated, Switch I and Switch II regions move inward and stabilize each other and the active state. Residues within the Switch III region form new salt bridges with Switch II residues in Gαt (1), which may aid in nucleotide binding. In Gαt, activation brings Trp207 in Switch II into a more hydrophobic environment, thus shielding tryptophan’s electrons and causing an increase in Gα intrinsic fluorescence. This is the basis for a widely used intrinsic Trp fluorescence assay for Gα activity (section 1.3.1). Upon activation with the GTP transition state analog, AlF4−, Gα-GDP-AlF4 adopts a conformation resembling the transition state for GTP hydrolysis, and this conformational change is accompanied by an increase in fluorescence of Trp207 in Gαt (Trp211 in Gαi). Crystal structures of Gt and Gi (4,6) reveal the plasticity of the Switch II region (disordered in GαiGDP), which adopts a distinct conformation in the presence of Gβγ. Switch II residues participate in binding to Gβγ. Gα subunits contact βγ dimers at the switch interface; changes in conformation of residues in this region occur during the activation process and result in the release of βγ, facilitating the interaction of Gα-GTP and βγ with distinct downstream effectors. Solved structures of inactive rhodopsin (7–11) have greatly advanced our understanding of G-protein-coupled receptors, coming some 10 years after the structures of heterotrimeric G-proteins were solved (1,2,5). The effort to obtain a high resolution structure of an activated GPCR is under intense investigation by a number of groups. A structure of an activated GPCR bound to G-protein is an even more daunting undertaking. While these efforts are ongoing, other biophysical methods can be used to elucidate details of receptor–G-protein interactions. Biophysical studies complement structural information, and there is still much to learn about conformational changes associated with GPCR signaling. Crystallographic studies are static by nature, whereas G-protein signaling is a dynamic process. Hence, the need exists for assays to probe conformational changes associated with signal transduction from an activated GPCR to G-protein in solution. Both fluorescence and electron paramagnetic resonance (EPR) are tools that can be used to examine changes in specific regions of G-proteins in solution. This can be achieved by attaching probes in a sitespecific manner and directly examining changes in the dynamics and environment of the probes as G-proteins undergo conformational changes during the G-protein cycle. These techniques are especially useful in examining changes in regions of the protein that are typically absent or disordered in crystal structures, such as the

7315.indb 4

11/19/07 8:25:36 PM

Biophysical Approaches to the Study of G-Protein Structure and Function

5

extreme amino and carboxy termini, both of which are known to interact with receptors and undergo activation-dependent changes. Biophysical studies of G-proteins have provided a wealth of information about dynamic changes that occur in G-protein signaling that cannot be gleaned from current crystal structures. Recent EPR studies reveal that the carboxy termini of G-proteins provide a critical link from receptor activation to GDP release through a rotation and translation of the α-5 helix (12). The α-5 helix in the carboxy region adjoins the β6/α5 loop that contains a TCAT motif involved in guanine nucleotide binding (3), coupling receptor activation to nucleotide release, which is the slow step in G-protein activation. Fluorescence and EPR studies have demonstrated that residues in the amino terminus of myristoylated Gαi proteins are relatively immobile and reside in a more solvent-excluded environment than the nonmyristoylated forms, consistent with an intramolecular binding site for the amino terminus of Gα proteins (13). Together with crystallographic information, these biophysical studies enlarge our understanding of the mechanisms underlying G-protein signaling.

1.1.1 Overall Strategy Because both fluorescence and EPR rely on a uniquely labeled protein, this requires creation of a parent protein lacking solvent-exposed cysteines, which can be accomplished with site-directed mutagenesis (SDM). Labeling of specific cysteine residues requires conservative replacement of solvent-exposed cysteines that would react with thiol directed labels. SDM can be performed easily, aided by numerous commercial kits available, to engineer specific labeling sites (in this case, cysteines) into a protein for dynamic fluorescence and EPR studies. This is done against a background lacking solvent-accessible cysteines to achieve labeling specificity. After the expression of the mutant protein containing cysteines at sites of interest, the protein is modified by fluorescent and spin labels, a dual strategy for the study of protein dynamics (scheme 1.1). Although EPR reveals changes in mobility and conformation at the backbone level as well as information regarding tertiary contacts, fluorescence can be used to gauge the overall environment of a specific residue. Together, these complementary approaches reveal information about the mobility and environment of a specific residue in regions undergoing conformational changes during G-protein signaling. Fluorescence has long been used to examine changes in protein dynamics and conformation. There is an ever-increasing variety of commercially available fluorescent probes, many with superior spectral characteristics. Advances in EPR technology allow distance measurements between labeled residues during the G-protein cycle. Although mutagenesis and expression of GPCRs are beyond the scope of this chapter, these techniques can be used in conjunction with similarly modified receptors to examine interactions between G-proteins and receptors during the G-protein cycle, using both fluorescence resonance energy transfer (FRET) and EPR double electron–electron resonance EPR spectroscopy (DEER). Once the parent protein lacking solvent-exposed cysteines (termed Hexa I) is generated, it becomes the background against which further mutations are made, allowing attachment of thiol-reactive labels at specific sites of interest. Expression and purification results in protein quantities sufficient for both fluorescent and EPR

7315.indb 5

11/19/07 8:25:37 PM

6


studies, and provides two independent and complementary measures of conformational changes that occur during the G-protein cycle. Although fluorescence reveals information about the relative hydrophobicity of the probe’s overall environment, EPR can resolve populations with distinct mobilities. For example, a highly dynamic region may demonstrate both a mobile and immobile component and the timescale of such motions (14). A mutational strategy that places these labels at a series of strategic points within the protein allows construction of an overall picture of protein conformational changes, as the labeled Gα proteins are interrogated throughout the G-protein cycle. Regions undergoing conformational changes during the cycle can be characterized by a gradient of dynamics, mobility, and environment using fluorescence and EPR approaches.

1.2 Protein Expression and Purification Gαi1 protein was selected as the basis of our studies of Gα proteins, because it is more amenable to expression in bacteria than Gαt and interacts nearly as well with rhodopsin (15). These two Gα subunit family members share an 80% sequence similarity (16), and the conformational changes observed in the Gαi1 background are likely to be a good approximation of those in Gαt. Prior to generating cysteine mutants at sites of interest for conformational studies, we created the Gαi Hexa I parent protein (15) by conservative replacement of solvent-exposed cysteines in Gαi protein. Native Gαil contains 10 cysteine residues (C3, C66, C139, C214, C224, C254, C286, C305, C325, and C351); however, only six of these are solvent- exposed, making them sensitive to thiol-reactive labels. We conservatively replaced these six cysteine residues (C3S, C66A, C214S, C305S, C325A, and C351I) using SDM, following the manufacturer’s protocol (QuickChange, Stratagene, La Jolla, California). The amino acids used to conservatively replace native cysteines were chosen based on the specific environment of the native cysteine, where serine was placed at more polar sites and alanine at less polar sites. As a result, most of the cysteines were replaced by serine, but metarhodopsin II stabilization assays of Gαi indicated that alanine was a better substitute for native serines at residues 66 and 325, whereas isoleucine was a better substitute for residue 351. The isoleucine substitution residue 351 was critical for maintaining normal interactions with the receptor, as serine and alanine substitutions prevented receptor-catalyzed GTPγS binding. Gαi Hexa I protein expresses well, is folded properly, and has rates of receptor-catalyzed (35S) GTPγS binding similar to wild-type Gαi1 (15). A hexahistidine tag was inserted between residues M119 and T120 of the Gαi1 coding region, which aided in purification of expressed protein without perturbing amino, carboxy, or GTPase domains, all of which undergo activation-dependent conformational changes (13,15,17,18). The site of the hexahistidine tag is located in a solvent-exposed region of the helical domain, far removed from these critical regions of the protein. The resulting Gαi Hexa I protein has properties functionally similar to that of Gαt (15), and provides the cysteine-less background against which later mutations were made. Following creation of the cys-less Hexa-I parent protein, selected individual residues in the parent Gαi Hexa I protein can be mutated to cysteine, for the purpose of attaching fluorescent or nitroxide spin labels at the selected site. After SDM to

7315.indb 6

11/19/07 8:25:37 PM


7

Gαi Hexa l

SDM Inserts Cys at Sites of Interest

Expression and Puriﬁcation

Fluorescence Analysis

Fluorescent Labeling

Nitroxide Spin Labeling

EPR Analysis

Scheme 1.1 Overall strategy.

specifically engineer in cysteine residues, proteins are expressed and assayed for function in the same manner as the Gαi Hexa I parent protein. This is then followed by labeling for biophysical studies to probe conformational changes at each position along a region of interest. An advantage of this approach is that expression and purification of each protein typically yields sufficient quantities for both fluorescent and EPR studies (scheme 1.1). Mutations that disrupt protein folding and function are often poorly expressed, therefore several adjacent residues in a region of interest may need to be mutated in order to build a picture of conformational changes within a region.

1.2.1 Expression and Purification of Gαi Hexa I Proteins BL21DE3 Gold E. coli (Stratagene) are transformed according to the manufacturer’s instruction with the expression vector containing the Gαil Hexa I cDNA (15), which was created by SDM of a Gαil expression vector (courtesy of M. Linder, Washington University, St. Louis, Missouri) with a hexahistidine coding region after residue 119 in the helical domain for ease of purification. Site-directed mutagenesis is commonly used to generate mutations, and the prevalence of commercially available kits and reagents allow for introduction of a mutation in one day. It is helpful to keep in mind that primers should be gel-purified, generally 25–45 bases in length, and with a melting temperature of at least 75–80°C. Best success is found with primers that consist of at least 40% GC content. Following manufacturer’s instructions, high-quality DNA is obtained that is suitable for sequencing and subsequent transformation into E. coli. The Gαil Hexa I parent protein and cysteine mutants for labeling experiments were expressed and purified according to the protocol below. Although a great many

7315.indb 7

11/19/07 8:25:38 PM

8


mutant proteins were generated (12,13,15), not all express well or result in a functional protein. Therefore, evaluation of proteins in functional assays is a critical part of the experimental plan. Note that Gαi Hexa I proteins can also be coexpressed with pbb131 plasmid (courtesy of M. Linder) encoding for N-myristoyl transferase in order to express proteins in their N-myristoylated form. The Gαi Hexa I vector encodes ampicillin resistance, whereas the NMT plasmid encodes kanamycin resistance, allowing for selection of expressed proteins in appropriately supplemented media. After sequence verification and transformation into E. coli, the protein is expressed and purified. As our system utilizes a hexahistadine tag, Ni-NTA purification is used to obtain a semipurified protein. An additional round of purification using anion-exchange chromatography is required in order to obtain purity of greater than 85–90%, suitable for labeling and functional assays. The protocol 1 generates quantities sufficient for functional assays and thiol-directed labeling, as summarized in scheme 1.1. Protocol 1: Expression and Purification of Gαi Hexa I Proteins

7315.indb 8

1. Transfected E. coli cells are grown to OD600 of 0.3 units in 2YT broth containing 100 µg/mL ampicillin. For myristoylated protein, 2YT broth is additionally supplemented with kanamycin, 50 µg/mL, and myristic acid, 50 µM. Typically, 50 mL overnight culture is then added to 500 mL of media, and grown to 0.5 OD600 over a period of 4–8 h. 2. Expression is induced with 30 µM isopropyl-β-d-thiogalactopyranoside at room temperature for 12–16 h with shaking. 3. Cell pellets from a 1 L culture are resuspended in 50 mM NaH2PO4, pH 8.0, 300 mM NaCl containing 5 mM imidazole and 1 µg/mL of the protease inhibitors pepstatin, leupeptin, and aprotinin. 4. Cells are lysed by sonication on ice, generally delivered in 30 s pulses with 1 min cooling time between pulses. Care should be taken to prevent sonicated extracts from excessive heat due to sonication. 5. The lysed extracts are subjected to 50,000 rpm centrifugation for a period of 1 h, and the supernatant from a 1 L culture is combined with 5 mL of a 50% slurry of NiNTA agarose (Qiagen, Valencia, California) and rotated at 4°C for 1 h to allow for complete binding to resin. 6. The slurry is loaded into a gravity filtration column, and flow-through is discarded. The remainder of unbound proteins is removed by washing with 30 mL of resuspension buffer (step 3), followed by an additional 30 mL wash with resuspension buffer containing 10 mM imidazole. 7. Gαi Hexa I proteins are eluted with 7 mL of resuspension buffer containing 40 mM imidazole. 8. The semipurified eluates are dialyzed overnight into a 1 L solution of 50 mM Tris, pH 8.0, 50 mM NaCl, 1 mM MgCl2, 20 µΜ GDP, 10 mM 2-mercaptoethanol, 100 µΜ PMSF, and 20% glycerol. 9. After dialysis, proteins are purified using anion exchange chromatography (MonoQ, Amersham, or other strong anion exchanger). Proteins are applied to column in a low-salt buffer consisting of 50 mM Tris, 50 mM NaCl,

11/19/07 8:25:38 PM


9

and 1 mM MgCl2 at pH 7.8. Purified protein is eluted in a linear gradient of 50–500 mM NaCl, with proteins eluting between 150–250 mM NaCl. Typically, 0.5 mL fractions are collected into tubes containing 300 µL of low-salt buffer supplemented with 50 µΜ GDP; including GDP in running buffer may obscure absorbance of Gα protein, as they both bind to anion exchange columns and elute at a similar salt concentration. 10. Purified protein fractions demonstrating at least 40% increase in intrinsic fluorescence upon AlF4− activation (see functional assays) are pooled, concentrated in a 10 kDa cut-off concentrator (Sartorius, Goettingen, Germany), and washed with low-salt buffer supplemented with 10 µΜ GDP. 11. Proteins are quantitated by the Coomassie blue method using bovine serum albumin as a standard, and further confirmed by Coomassie blue staining of SDS-PAGE gels, migration consistent with a molecular weight of 37 kDa (figure 1.1). 12. Proteins are concentrated and stored in low-salt buffer supplemented with 10 µΜ GDP and 10% glycerol prior to storage at −80oC. For Gαi proteins containing native, solvent-exposed cysteines, 1 mM DTT is included in the storage buffer, and must be completely removed prior to labeling with any thiol-reactive reagents by extensive buffer exchange or use of P6 (Biorad) desalting gel. Gαi Hexa I parent protein and Gαi Hexa I proteins with cysteine mutations at the sites of interest (hereafter referred to as Hexa I proteins) are ready for labeling if stored in the absence of DTT or BME. Alternatively, proteins for storage can be supplemented with Tris(2-carboxyethyl) phosphine hydrochloride (TCEP), a reducing

37 kDa

Figure 1.1 SDS-PAGE of Gαil Hexa I protein after anion exchange purification, 5 µg, 10–20% Tris-Glycine gel, visualized by Coomassie blue staining.

7315.indb 9

11/19/07 8:25:39 PM

10


agent that is reported to be compatible with thiol-reactive reagents, eliminating the need for their removal prior to the labeling process. However, the presence of TCEP may also reduce labeling efficiency, so one may need to consider this carefully and optimize labeling conditions in the presence of this reducing agent (section 1.4.1). In Hexa I proteins, substitution of native solvent-exposed cysteines considerably reduces the propensity for dimerization between adjacent molecules.

1.2.2 Rhodopsin and Gβγ Preparation Rhodopsin and Gβ1γ1 (Gβγ) are obtained from native retinas as detailed previously in Reference 18. Unwashed rod outer segments (ROS) are stored at −80°C and in buffer containing 10 mM MOPS, pH 8.0, 90 mM KCl, 30 mM NaCl, 2 mM MgCl2, 0.1 mM EDTA, 1 mM DTT, and 50 μM PMSF. To obtain rhodopsin free of endogenous G-proteins, ROS membranes were stripped of G-proteins with 4 mM urea as described in (19). Rhodopsin concentration was determined by measuring the absorbance of solubilized ROS membrane suspensions at 500 nm before and after photobleaching (protocol 2). Gβγt was prepared from ROS, as summarized in protocol 3 below. Protocol 2: Measurement of Rhodopsin Concentrations in Urea-Washed ROS Samples

1. Because quantitation of rhodopsin requires comparison of absorbances between dark- and light-activated rhodopsin, the assay must be performed under dim red light, using a filter that passes light greater than 650 nm. A water-jacketed cuvette holder is used to keep the temperature of the cuvette between 5–10°C. 2. Under dim red light (Kodak, Filter 1), 25 µL of rhodopsin is added to 500 µL of 20 mM hexadecyltrimethyl ammonium chloride, mixed well and placed in the cuvette chamber. Absorbances are recorded at 500 nm and 650 nm (reference). 3. The sample is then illuminated with three successive flashes of bright white light. 4. For each dark and light measurement at 500 nm, subtract the absorbance of reference measured at 650 nm. To calculate rhodopsin concentration from Beer’s law, the difference in absorbances between light-activated rhodopsin measurement (less 650 absorbance) and dark-adapted rhodopsin (less 650 absorbance) is calculated, and multiplied by the dilution factor used. With the extinction coefficient (ε) of 42,000 cm−1M−1 for rhodopsin, and a path length specific for the instrument (L), concentration is calculated from absorbance, where absorbance = ε C L. Protocol 3: Gβ1γ1 Preparation

First, transducin is extracted from ROS, followed by chromatography to isolate Gβγ from Gα as described in detail in Mazzoni et al. (18). Briefly, the process is:

7315.indb 10

11/19/07 8:25:39 PM


11

1. ROS membranes are washed four times with isotonic buffer (5 mM Tris, pH 8.0, 120 mM KCl, 0.6 mM MgCl2, 1 mM DTT, 0.1 mM EDTA, and 0.1 mM PMSF) and twice with hypotonic buffer lacking KCl. 2. Transducin is eluted from the membranes by washing twice with hypotonic buffer containing 0.1 mM GTP. 3. To isolate Gβγ from transducin, transducin is applied to a HiTrap™ Blue HP blue sepharose column (Amersham Biosciences, Piscataway, New Jersey) at 4°C using a peristaltic pump in 10 mM Tris, pH 7.5, 150 mM NaCl, 20 mM MgSO4, 1 mM EDTA, 10% glycerol, and 14.3 mM 2-mercaptoethanol. 4. MgSO4 encourages subunit dissociation, and as Gβγ does not bind the blue Sepharose column, it is obtained in the flow-through fraction, whereas Gαt remains bound to the column. 5. The flow-through fractions from chromatography are concentrated and evaluated on SDS-PAGE to identify fractions containing pure Gβγ, which can be pooled and stored in buffer containing 10 mM Tris, pH 7.5, 100 mM NaCl, 5 mM 2-mercaptoethanol, and 10% glycerol at −80°C. 6. Just prior to use, Gβ1γ1 samples are buffer exchanged into 20 mM MES, pH 6.8, 100 mM NaCl, 2 mM MgCl2, and 10% glycerol.

1.3 Functional Assays 1.3.1 Intrinsic Tryptophan Fluorescence A wide variety of functional assays exist to examine functional integrity of Gα proteins. The intrinsic tryptophan fluorescence assay demonstrates the ability of Gα proteins to undergo an activation-dependent conformational change in solution. In this assay, aluminum fluoride (AlF4−) complexes with bound GDP to form a transition state that mimics the conformation of Gα during the transition state associated with GTP hydrolysis. A tryptophan residue within the Switch II region acts as a reporter of this activation-dependent conformational change. This conformational change is reflected by an increase in tryptophan fluorescence emission for properly folded, active Gα proteins. The increase in emission is typically expressed relative to its emission in the GDP-bound state; minimally, a 40% increase over basal is exhibited by functional Gα subunits. This assay allows efficient screening of fractions for activity after HPLC purification. Oftentimes, only selected fractions from the peak eluting at retention times that correspond to Gα has functional activity, and it is necessary to discard inactive fractions prior to pooling and concentrating the purified Gα protein. Protocol 4: Intrinsic Tryptophan Fluorescence

7315.indb 11

1. Using a fluorescence spectrophotometer, set excitation/emission wavelengths to 280/340 nm. 2. Zero emission reading from buffer consisting of 50 mM Tris, 50 mM NaCl, 1 mM MgCl2, and 10 µΜ GDP. 3. Add 100 nM of Gα subunit, mix well, and measure emission in the GDPbound state.

11/19/07 8:25:40 PM

12


4. To form activated GαGDP-AlF4− complexes, add NaF to 10 mM and AlCl3 to 50 µΜ, and mix. 5. Measure emission after addition of AlF4−; for properly folded, active Gα subunits, the emission should be at least 40% greater in the GDP-AlF4−bound state compared to the Gα-GDP-bound form (figure 1.2a).

1.3.2 Basal Nucleotide Exchange G-proteins undergo GDP release under basal conditions at a relatively low level, as compared to rates in the presence of activated receptor. Furthermore, this rate is (a)

25

+ Gαil

Fluorescence, A.U.

20 15 10 + AlF –4

5 0 –5

(b)

0.0

0.5

1.0

1.5 2.0 Time (min)

2.5

3.0

2.5

3.0

+ Gαi Hexa l

Fluorescence, A.U.

25 20 15 10

–

+ AlF 4

5 0 –5

0.0

0.5

1.0

1.5 2.0 Time (min)

Figure 1.2 Intrinsic tryptophan activation of Gα proteins. Gαil (a) or Gαil Hexa I (b), 100 nM, is added to buffer consisting of 50 mM Tris, 50 mM NaCl, 1 mM MgCl2, and 10 µΜ GDP. Tryptophan emission is monitored at excitation/emission 280/340 nm, both before and after addition of NaF to 10 mM and AlCl3 to 50 µΜ. Proteins exhibit ≥ 40% increase in fluorescence in GDP-AlF4−-activated form, relative to inactive, GDP-bound form, as would be expected for properly folded and fully functional Gα proteins.

7315.indb 12

11/19/07 8:25:45 PM

13

Biophysical Approaches to the Study of G-Protein Structure and Function H3C O N N B H3C

F

F

N

NH2

N CH2NH

O C

CH2

S

O P O O–

O P O O–

O P OCH2 O–

HO

N O

N

OH

Figure 1.3 Structure of BD-GTPγS.

somewhat increased for Gαi relative to Gαt. The low rate of intrinsic GDP exchange for Gαt subunits supports fast, efficient signal termination required for these subunits during the visual transduction process. In order to measure intrinsic nucleotide exchange in the absence of a receptor, a fluorescently modified GTPγS can be used to monitor GDP release (protocol 5), as can intrinsic tryptophan fluorescence (protocol 6). Both of these methods avoid the use of radioactivity, and results from these assays correlate well with published radioactive assays for Gαi subunits. Bodipy FL-GTPγS (Invitrogen, Madison, Winconsin), has a fluorescent label on the γ-phosphate of GTPγS (figure 1.3). The emission of Bodipy GTPγS is quenched in aqueous solution, and increases upon binding to a number of GTP-binding proteins, including Gαi proteins (20). Bodipy-GTPγS provides a convenient tool to measure GDP release. As the spectral properties of the probe overlap with rhodopsin, this GTP analog is more suited to intrinsic nucleotide exchange assays than receptor-mediated assays. Intrinsic tryptophan fluorescence can also be used to monitor GTPγS binding (both basal and receptor-stimulated), increasing options for nonradioactive monitoring of nucleotide binding. Protocol 5: Basal Nucleotide Exchange Assay by Extrinsic Fluorescence

7315.indb 13

1. Emission of 5 µΜ Bodipy-GTPγS (BD-GTPγS, figure 1.3) is monitored for approximately 1 min at excitation/emission 490/512 nm with stirring at 18°C in 50 mM Tris, 50 mM NaCl, and 1 mM MgCl2 at pH 7.8. 2. 500 nM Gαi (with or without Gβγ in a 2 molar excess) is added to the cuvette with BD-GTPγS, and emission is monitored over a 5 min period (figure 1.4a) using a Varian Cary Eclipse (Mulgrave, Australia).

11/19/07 8:25:45 PM

14

Signal Transduction in the Retina (a)

Basal Nucleotide Exchange

15.0

Relative Fluorescence [(F – Fo)/Fo)]*100

12.5 10.0 +Gαil

7.5 5.0 2.5 0.0 –2.5

0

1

2

3

4

5

6

7

Time (min) (b)

Basal Nucleotide Exchange

15.0

Relative Fluorescence [(F – Fo)/Fo)]*100

12.5 10.0 7.5

+ Gαβγil

5.0 2.5 0.0 –2.5

0

1

2 3 Time (min)

4

5

Figure 1.4 Basal nucleotide exchange of Gαil subunits, measured in the presence of 5 µΜ BD-GTPγS in buffer consisting of 50 mM Tris, 50 mM NaCl, 1 mM MgCl2, and pH 7.8. Excitation/emission set at 490/512 nm using a Varian Eclipse fluorescence spectrophotometer. (a) 550 nM Gα or (b) reconstituted heterotrimer formed from preincubation of 550 nM Gα and 1 µΜ Gβγ for 5 min at 4°C prior to addition to a cuvette containing 5 µΜ BD-GTPγS. Data are normalized to basal BD-GTPγS emission measured in the absence of G-protein, and represented as percent increase over basal. Basal nucleotide exchange can also be measured by monitoring conformational changes in the Switch II region by monitoring Trp emission, as described in the intrinsic fluorescence assay (protocol 4).

Protocol 6: Basal Nucleotide Exchange Assay Using Intrinsic Tryptophan Fluorescence

7315.indb 14

1. Using a fluorescence spectrophotometer, set excitation/emission wavelengths to 300/345 nm. 2. Zero the emission on buffer alone (10 mM MOPS, pH 7.2, 150 mM NaCl, and 2 mM MgCl2) at 15°C. 3. Add 500 nM Gα subunit to cuvette, mix, and monitor emission for 1 min.

11/19/07 8:25:47 PM


15

60

Intrinsic Fluorescence (% Max)

50 40 30 20 10 0 –10 –250

0

250

500

750

1000 1250 1500 1750 2000 2250 Time (s)

Figure 1.5 Increase in the intrinsic fluorescence of Gαi1 upon the addition of GTPγS (t = 0). Data are the average ±SEM of three independent experiments.

4. Add GTPγS to 10 µΜ and continuously monitor the emission over a period of 40 min (figure 1.5). Add NaF (10 mM) and AlCl3 (50 µΜ) to fully activate the subunits at the end of 40 min.

1.3.3 Receptor-Mediated GTPγS Binding Although radiolabeled GTPγS-binding assays have long been used in determination of GTPγS binding to Gα subunits, nonradiative methods are often preferable when available. Using nonradioactive nonhydrolyzable GTP analogs (protocols 7 and 8), receptor-mediated GTP binding of Hexa I proteins can be conveniently measured in a cuvette or 96-well plate format. The 96-well format allows for screening a large number of recombinant proteins quickly, with minimum sample volumes required. Protocol 7: Receptor-Mediated GTPγS Binding Using Extrinsic Fluorescence

7315.indb 15

1. The Eu-GTP binding assay is based on the DELFIA Europium-GTP binding kit (PerkinElmer) using a Victor fluorescence spectrophotometer (PerkinElmer). 2. Under dim red light, incubate 1 nmol rhodopsin with 1 nmol G-protein in a 2 mL microcentrifuge tube for 10 min on ice. 3. Add 1.8 mL of buffer containing 100 mM HEPES (pH 8.0), 1 mM EDTA, 10 mM MgSO4, and 10 mM DTT and mix well to resuspend the membrane. 4. Wash each well of a Pall-Gelman AcroWell 96-well filter plate with 0.45 µm GHP nitrocellulose membrane once with 300 µL cold dilute GTP Wash Buffer from the kit.

11/19/07 8:25:48 PM

16


5. Photolyze rhodopsin and add 200 µL of 100 nM Eu-GTP (10 nM final concentration) to initiate the reaction. 6. Filter 2 × 100 µL reaction at each time point (1, 3, 5, 7, 10, 15, 20, and 25 min). Wash with 3 × 300 µL cold GTP Wash Buffer. 7. After the final time point, disconnect the vacuum and add 50 µL of GTP Wash Buffer to each well. 8. Generate a Eu-GTP standard curve by adding 10 µL of serial dilutions from 0–5 nM to the filter plate. Allow dilutions to absorb into the filter, then add 50 µL of cold GTP Wash Buffer. 9. Read the plate in the Victor fluorescence spectrophotometer using the manufacturer’s preset protocol for Europium time-resolved fluorometry. 10. Convert counts to moles of Eu-GTP-bound using the standard curve (figure 1.6). Protocol 8: Receptor-Mediated GTPγS Binding Using Intrinsic Fluorescence

1. Using a Varian Cary Eclipse (or any other standard fluorescence spectrophotometer), zero the instrument on buffer alone containing 10 mM MOPS, pH 7.2, 150 mM NaCl, 2 mM MgCl2 at 15°C, using excitation/emission wavelengths of 300/340 nm. 2. In a separate tube, reconstitute heterotrimer by incubation of 500 nM Gα and 500 nM Gβγ on ice for 15 min prior to assay. 3. Add heterotrimer plus 100 nM rhodopsin to the cuvette and monitor the emission over 1 min to obtain baseline. 4. Add GTPγS to 10 µΜ and monitor the emission over 40 min at 15°C. 5. Normalize data to the baseline (set to 0%) and the maximum fluorescence at 100%, set to the value obtained by activation by AlF4−. Eu-GTP-Bound G Protein (fmol)

30

Dark Light

20

10

0

0

10

20

30

Time (min)

Figure 1.6 Basal (closed circles) and rhodopsin-catalyzed (open circles) binding of EuGTP to transducin under conditions described in protocol 7. Data are the average ±SEM of three independent experiments.

7315.indb 16

11/19/07 8:25:49 PM


17

6. The rate of receptor catalyzed nucleotide exchange can be calculated by fitting the data to the equation F = Fmax(1−e−kt), where F is the percent of maximum fluorescence (Fmax) at time t s and k is the catalytic activation rate constant for GTPγS binding in s−1.

1.3.4 Receptor Binding G-proteins couple to activated receptors, which catalyze release of GDP and uptake of GTP, and both GαGTP and Gβγ then act on distinct downstream effectors. A traditional rhodopsin-binding assay is used to confirm the ability of Hexa I proteins to bind to rhodopsin, a prerequisite to measuring receptor-catalyzed changes in the Hexa I proteins using EPR and fluorescence methods. In the rhodopsin binding assay (protocol 9), Gα is incubated with Gβγ and rhodopsin under three conditions: dark, light, and light with GTPγS. The samples are centrifuged, and the membrane and supernatant fraction are analyzed by SDS-PAGE. Upon light activation, the G-protein moves from the supernatant to the membrane, and it is released upon the addition of GTPγS (figure 1.7). Protocol 9: Rhodopsin-Binding Assay

1. Incubate 10 µΜ Gα hexa I protein with an equimolar amount of Gβγ at 4°C for 15 min. 2. Incubate urea-washed ROS membranes (100 µΜ) with 10 µΜ reconstituted heterotrimer at 20°C with 100 µΜ GTPγS for 30 min under light conditions. 3. Following this 30 min incubation, pellet membranes by centrifugation at 20,000 × g for 1 h. 4. Pellet and supernatant fractions are then analyzed by SDS-PAGE (figure 1.7), visualized with Coomassie blue and quantified by densitometry (Molecular Imager ChemiDoc™, XRS System, Biorad, Hercules, California). 5. The percent bound can be calculated from the equation below, where GαL and GαG are the amount of Gα in the light and GTPγS supernatants, respectively.

% Bound = 1 − (GαL/GαG) Dark P

Light S

P

GTPγS S

P

S Gαi1 Gβ1

Figure 1.7 Rhodopsin-binding assay. Gα is incubated with Gβγ and rhodopsin in dark (left), light (center), and light with addition of GTPγS (right). Pellet (P) versus supernatant (S) are analyzed by SDS-PAGE.

7315.indb 17

11/19/07 8:25:50 PM

18


1.4 Fluorescent probes for analysis of conformational changes in GPCR signaling The engineering of cysteine residues into sites of interest (in the background of the Gαi Hexa I parent protein with solvent-exposed cysteines removed) provides a valuable tool to study changes in specific regions of Gα proteins during the activation cycle by both fluorescence and EPR. Understanding the data obtained from these two techniques together provides a fuller, more comprehensive picture of protein dynamics than either technique alone. There is a great variety of commercially available fluorescent probes to choose from when labeling Gα subunits. Some probes such as bimane and 1-anilino-8-naphthalene sulfonate (ANS) exhibit environmentally dependent spectral shifts, whereas other probes report changes in environment by changes in emission intensity. It is important to choose a probe or probes that reflect conformational changes without perturbing protein function and do not conflict with spectral characteristics of other components used in the assay. For example, rhodopsin absorption maxima change as it moves through the G-protein cycle. As probes can be charged or uncharged, hydrophobic or hydrophilic, it is important to consider their potential effects on the labeled protein and its ability to interact with other proteins. To control for this, conformational changes measured by fluorescence are best measured with several probes of varying size and chemistry. This is generally not difficult to achieve, given the variety of fluorescent probes currently available (Invitrogen, Madison, Wisconsin; Toronto Research Chemicals, Toronto, Canada).

1.4.1 Fluorescent Labeling of Gα Subunits In this section we describe various methods to label Gα subunits for fluorescence assays. First, one must choose a suitable fluorescent label or labels. For example, if one is conducting a protein–protein interaction assay with labeled Gα subunit as reporter of the interaction, Lucifer yellow and bimane are good choices, as they are relatively small in size and moderately water soluble. Bodipy, although somewhat more hydrophobic, has a higher quantum yield, making it a sensitive reporter of changes in emission intensity. Bimane and other probes demonstrate shifts in their emission maxima as a result of changes in conformation or binding that exclude solvent from the probe’s environment. If the labeled proteins will be used in conjunction with rhodopsin, then probes with spectral characteristics in the red region are preferred in order to avoid spectral interferences from rhodopsin. It is also important to ensure the probe is photostable (resists bleaching) and does not perturb protein function. Therefore, all labeling should be accompanied by functional assays to ensure that the functional integrity of the labeled protein is maintained. The low degree of background labeling in the Gαi Hexa I parent protein promotes labeling at the cysteine residue of interest. However, cysteine mutations in regions of the protein with reduced solvent accessibility may result in poor labeling, despite good functional activity of the unmodified protein; therefore, labeling efficiency at different locations within the protein may vary considerably. Labeling conditions may need to be optimized for specific cysteine mutants. For example,

7315.indb 18

11/19/07 8:25:50 PM


19

labeling at the extreme termini of the protein is very efficient, requiring less labeling time than other regions of the protein. Labeling efficiency also depends on the nature of the probe used; probes of a more hydrophobic nature, which exhibit poor solubility in water, tend to label less efficiently than water-soluble probes. This can be ameliorated to some extent by labeling in the presence of DMSO, which can be used at varying concentrations up to 25% for short periods without adversely affecting Gαi protein’s ability to undergo activation-dependent conformational changes. Because only labeled residues act as reporters in these fluorescence assays, a relatively modest degree of labeling is generally sufficient (7 mM) and 1 mL each of 3 mg/mL DNase I, DNase II, and RNase (Sigma) to each bottle, mix thoroughly, and incubate on ice for 40 min. Transfer the suspension to 250 mL bottles and pellet cell debris in a GSA rotor at 9000 rpm for 90 min. Transfer supernatant (~600 mL) to a 1 L beaker. Under continuous gentle stirring at 4°C, add 192 g (to 0.32 g/mL) of ammonium sulfate in 4–5 portions over a 20–40 min period. Let stirring continue until all ammonium sulfate is dissolved, and then pellet the precipitated protein in a GSA rotor at 9000 rpm for 90 min. Carefully remove supernatant and floating white material, and wipe the walls clean. Cover the tube with foil and store at −80°C until needed (stable for at least a month). 4. Heparin-Sepharose chromatography. Dissolve the protein in 200 mL of ice-cold column buffer (10 mM Tris-HCl, pH 7.5, 2 mM EDTA, 2 mM EGTA, 2 mM benzamidine, 1 mM PMSF) (CB). It takes 20–40 min of gentle shaking (avoid frothing). Pellet the insoluble material (GSA rotor, 9000 rpm, 90 min), and carefully determine the volume of the supernatant to estimate the volume of the initial pellet (it is usually 12–18 mL and contains 1200–1600 mg of total protein). Filter the supernatant through a 0.8 µm Millipore filter. Dilute the filtrate with CB to 65 times the volume of the initial pellet for wild-type visual arrestin during loading onto a 25 mL column of Heparin-Sepharose, equilibrated with CB/0.1 M NaCl using a gradient mixer and an appropriate ratio of filtrate and CB. Load at 1.5–2.5 mL/min, wash the column with ~200 mL of CB/0.1 M NaCl. Elute with a 400 mL linear gradient (CB/0.1 M NaCl -> CB/0.4 M NaCl) and collect 10 mL fractions. The main rod arrestin peak elutes in 8–12 fractions between 200–240 mM NaCl. 5. Q-Sepharose chromatography: Pool arrestin-containing fractions (total protein 15–30 mg, ~50% pure arrestin), concentrate to ~10 mL (Amicon, YM-30 membrane), and filter through a 0.8 µm Millipore filter. Calculate the concentration of NaCl in this pool, and load the sample at 1 mL/min onto a 10 mL Q-Sepharose column (equilibrated with CB) while diluting

11/19/07 8:33:19 PM

64


the sample (using a gradient mixer) with CB to a final NaCl concentration of ~10 mM. It is important that the protein be diluted just before it is pumped onto the column because arrestins aggregate at low salt, but survive it while bound to the resin. Wash the column with 40–80 mL of CB, and then elute with a 400 mL linear gradient (CB -> CB/0.1 M NaCl). Arrestin (>95% purity) elutes in a peak at about 60 mM NaCl. Pool arrestin-containing fractions, add NaCl to 100 mM, and concentrate to 0.5–3 mg/mL (expect 2–5 mg/L of bacterial culture), filter through a 0.8 µm Millipore filter, aliquot, and freeze at −80°C. Frozen arrestins are stable for >> 2 years, and 2–3 freeze–thaw cycles do not appreciably reduce their activity. Protocol 5: Direct Binding Assay for Arrestin–Microtubule Interactions We developed two alternative methods, using either radiolabeled in vitro translated arrestin or purified protein. Expression of radiolabeled arrestins in cell-free translation is simple and easy, allowing the production of up to 20 different mutants suitable for functional analysis in 1 day (35,37). However, the level of nonspecific “binding” in this case is relatively high, up to 30% of specific binding. The assay with purified arrestin gives an excellent signalto-noise ratio, but requires prior protein purification and involves much more labor-intensive quantification by Western blot. Microtubule preparation from purified tubulin (Cytoskeleton, Inc.) should be performed according to the manufacturer’s instructions. Briefly, purified tubulin (>99%) is incubated at a concentration of 5 mg/mL in a buffer containing 80 mM PIPES, pH 6.9, 1 mM MgCl2, 1 mM EGTA, 10% glycerol, and 1 mM GTP at 37°C for 40 min. Taxol (20 µM) is added, and the tubulin is incubated at room temperature for 10 min. The polymerized microtubules are then diluted to a final concentration of 1 mg/mL in 80 mM PIPES, pH 6.9, 1 mM MgCl2, 1 mM EGTA, and 20 µM taxol. Version 1—Using radiolabeled arrestin produced in cell-free translation (35,38). This assay is similar to the assay we use for direct receptor binding, which has been previously described in detail (35). The separation of microtubule-bound and free arrestin by centrifugation is based on the huge difference in size between polymerized MTs and free arrestin (29). It is important to keep the samples warm (>25°C) throughout the procedure to prevent depolymerization of tubulin.

7315.indb 64

1. Incubate 200 fmol of radiolabeled in vitro translated arrestin in 100 µL of 50 mM Tris-HCl, pH 7.4, 0.5 mM MgCl2, 1.5 mM dithiothreitol, 1 mM EGTA, and 50 mM potassium acetate for 20 min at 25°C with 20 µg of prepolymerized tubulin. 2. Pellet microtubules along with bound arrestin by centrifugation at 90,000 rpm in a Beckman TLA 120.1 rotor for 10 min at 25°C. MT-arrestin pellets cannot be washed because of the low affinity of the interaction (i.e., high dissociation rate; see following text).

11/19/07 8:33:19 PM

How Rod Arrestin Achieved Perfection

65

3. Dissolve the pellet in 0.1 mL of 1% SDS, 50 mM NaOH, transfer to scintillation vials, add 5 mL of water-miscible scintillation fluid (we use ScintiSafe Econo2, Fisher), and quantify bound arrestin by liquid scintillation counting. 4. Nonspecific “binding” (arrestin pelleted in the absence of microtubules) should be determined in the same experiment and subtracted. Version 2—Using purified arrestin. We express wild-type rod arrestin and its various mutant forms in E. coli and purify it, as described in Protocol 4 and elsewhere (35), with yields ranging from 2 to 5 mg/L of bacterial culture. This method with minor modifications can be used to obtain most members of the arrestin family with purity suitable for crystallization (32,33,39).

1. Incubate 4 µg of purified arrestin in 50 µL of 50 mM Tris-HCl, pH 7.4, 100 mM NaCl, 2 mM EDTA, and 1 mM EGTA for 30 min at 30°C with 40 µg of prepolymerized tubulin. 2. Pellet microtubules along with bound arrestin by centrifugation at 90,000 rpm in a Beckman TLA 120.1 rotor for 10 min at 30°C. Parallel samples with the same amount of arrestin without microtubules serve as controls. 3. Dissolve pellet in SDS sample buffer. Because arrestin runs too close to tubulin on SDS-PAGE, bound arrestin must be quantified by Western blot (running different known aliquots of the same arrestin protein in parallel on each gel to construct a calibration curve, as described in detail in protocol 8).

3.4 Rod Arrestin Self-Association: The Shape and Function of the Tetramer Curiously, self-association of visual arrestin was discovered many years ago, when this protein was commonly known as “S-antigen” and its function remained obscure (40). It was recognized as a special feature of the visual system after the finding that rod arrestin crystallizes as a tetramer (dimer of dimers) (33,41), in sharp contrast to arrestin2 (32,42) and cone arrestin (39). Subsequent studies using sedimentation equilibrium (34) and small-angle x-ray scattering (43,44) to investigate arrestin selfassociation in solution yielded contradictory results: arrestin was found to preferentially form either a dimer (34,44) or a tetramer (43). More importantly, these studies did not experimentally address the structure of the arrestin oligomer in solution and its function in rod photoreceptors. A recent study comprehensively reexamined the mechanism of rod arrestin self-association using multiangle laser light scattering to determine the weight-average molecular weight (MWav) as a function of total monomer concentration (45). The advantages of this method include high resolution to within a few hundred Daltons, wide molecular mass range, relatively small sample size, and high sample throughput. Importantly, because the wavelength of light is large compared to the dimensions of proteins, the scattering is independent of molecular shape (46), making the interpretation of the data independent of speculative assumptions. In this recent analysis, the working range of arrestin concentrations was significantly expanded (up to 100 µM), which resulted in experimental detection of molecu-

7315.indb 65

11/19/07 8:33:20 PM

66


lar masses far in excess of that of arrestin dimer (90 kDa), proving the existence of tetramer in solution (45). These data confirmed the model: 2M = D (K1), 2D = T (K2), where M, D, and T are monomer, dimer, and tetramer, respectively (MDT model). The wide range of concentrations used (1–100 µM) greatly increased the precision of the measurements, allowing reliable determination of monomer–monomer (dimerization) and dimer–dimer (tetramer formation) association constants, K1 and K2. The corresponding dissociation constants (K D1 = 1/K1; K D2 = 1/K2; expression in molar units makes these more informative for a biologist) were found to be 37.2 and 7.4 µM, respectively (45). Thus, arrestin oligomerization is cooperative in the sense that the association constant K2 > K1 (K D1 > K D2), i.e., dimer formation is rate limiting, so that as soon as the dimer is formed, it associates into a tetramer. Based on the values of K1 and K2, tetramer is expected to be the dominant form at physiological arrestin concentrations in rod photoreceptors (>> 200 µM, based on the arrestin-rhodopsin expression ratio of 0.8:1 in mouse rods that has been determined independently by two groups (19,20). Protocol 6: Analysis of Arrestin SelfAssociation by Light Scattering This is the most direct approach for studying the self-association of arrestin (or any other protein, for that matter). The main advantage is that the interpretation of light scattering data does not require any assumptions regarding the molecular shape of the complex. This turned out to be particularly advantageous for rod arrestin because the shape of the biologically relevant solution tetramer at physiological pH and ionic strength was found to be dramatically different from that of the crystal tetramer (45), even though essentially the same tetramer was observed in crystals grown under different conditions (33,41). This method is fairly fast (one measurement takes 10–15 min), allowing one to perform multiple experimental measurements within a few hours. The only drawback of this method is that for a comprehensive series of measurements one needs fairly large amounts (>5 mg) of highly purified arrestin (45), although still not as much as for analytical centrifugation (34). The following procedure described was adapted from Reference 45. The design and construction of a fully functional cysteine-less base mutant of visual arrestin that was necessary for complementary site-directed spin labeling and electron resonance studies has been described in sufficient detail previously, so the reader is referred to the original papers (29,36,45).

7315.indb 66

1. Thaw purified wild-type or mutant arrestin (protocol 4) and centrifuge at >100,000 × g for 60 min at 4°C to remove any aggregated material. 2. Make light scattering measurements with a DAWN EOS detector coupled to an Optilab refractometer (Wyatt Technologies) following gel filtration on a 7.8 mm (ID) × 15.0 cm (L) QC-PAK GFC 300 column (Tosoh Bioscience). Load the arrestin samples (100 µL) at different concentrations at 25°C, at a flow rate of 0.8 mL/min in 50 mM MOPS, 100 mM NaCl, pH 7.2. Note that this column does not resolve oligomeric species, but simply acts as a filter to

11/19/07 8:33:20 PM

67


remove highly scattering particulates. Measure light scattering at 18 angles (15°–160°), absorbance at 280 nm, and refractive index (at 690 nm) for each sample for a narrow slice at the peak of the elution profile (47). 3. Calculations: The relationships describing the arrestin equilibria according to the MDT model are: K1 =

D M2

K2 =

T D2

where: M, D, and T are the monomer, dimer, and tetramer concentrations, respectively, and C is the total protein concentration. From these relationships and particular values for K1, K2, and C, the concentrations of monomer, dimer, and tetramer may be obtained as a solution to equations (3.1)–(3.3):

M + 2 ⋅ K1 ⋅ M2 + 4 ⋅ K12 ⋅ K 2 ⋅ M 4 − C = 0

(3.1)

D = K1 ⋅ M 2

(3.2)

T = D 2 ⋅ K 2 = K12 ⋅ K 2 ⋅ M 4

(3.3)

Using the values of M, D, and T, defined earlier, compute MWav (average molecular weight measured by light scattering) as:

(

)

MWav = Mm ( M + 4 D + 16T) / ( M + 2D + 4 T)

(3.4)

where Mm is the monomer molecular weight (45,000). Determine the value for C in the light scattering cell at the point where data were collected from the refractive index increment (0.184 g/mL) and A280 (both of which should yield the same value within experimental error). The experimental Mav values are then obtained from the concentration and light scattering data using Astra for Windows 4.90 software (Wyatt Technologies). Fit the experimental data for MWav as a function of C to equations 3.1–3.4 with only K1 and K2 as adjustable parameters using a least-squares method. Collecting the data at more than 10 experimental points ensures high precision of determination of arrestin self-association constants. In all published crystals of rod arrestin (33,41) grown in different conditions, the asymmetric unit contains four arrestin molecules arranged such that one pair is related by a local twofold rotation axis to a second pair, i.e., essentially the same tetramer (dimer of dimers). Therefore, all subsequent studies of arrestin self-association were based on the assumption that the solution tetramer had the same shape and structure (34,43,44). From a functional point of view, the main problem with this assumption was that based on the accessibility of two of the four monomers, it was impossible to explain why arrestin does not form “infinite” chains consisting of these tetramers, i.e., why self-association stops at the tetramer level. A recent study using arrestins spin-labeled in different positions finally solved this problem

7315.indb 67

11/19/07 8:33:21 PM

68


presenting several lines of compelling evidence that the shape of the physiologically relevant solution tetramer is quite different (45). First, numerous mutations in the “crystal” interfaces did not affect arrestin self-association, whereas others far from any “crystal” intersubunit interfaces severely interfered with it. Second, the pattern of spin label immobilization due to tetramer formation did not match the changes that would be expected if the crystal interfaces were involved. Finally, the great majority of long-range distance measurements (using double electron–electron resonance) made between the spin labels in different monomers within the tetramer were completely at odds with the distances expected in the crystal tetramer (45). This study established that in solution visual arrestin forms “closed” diamond-shaped tetramers, which explains both the cooperativity of arrestin self-association and the absence of higher-order oligomers (45). Most important from the biological viewpoint, long-range distance measurements that report the self-association status of visual arrestin also made testing the functional capabilities of the tetramer possible for the first time. It was shown that the addition of light-activated phosphorhodopsin (in an amount sufficient to bind the arrestin present in the assay) resulted in complete disappearance of all distances characteristic of arrestin oligomers, unambiguously demonstrating that only arrestin monomer can bind rhodopsin (45). In fact, in the solution tetramer a large part of the receptor-binding surface of arrestin that was previously mapped using a variety of methods (36,48–50), is engaged by “sister” monomers, thus providing a simple mechanistic explanation for its inability to bind rhodopsin. These data support the idea that the arrestin oligomer is a “storage” form, supplying active binding-competent monomer, as needed. Unexpectedly, the addition of microtubules at a concentration high enough to bind most of the arrestin did not affect the distances, indicating that arrestin dimer and tetramer are capable of interacting with microtubules (45). Because microtubules engage essentially the same surface in arrestin monomer as the receptor (29), the interaction of the tetramer with MTs must be mechanistically different. It is tempting to speculate that the ability of arrestin tetramer to bind microtubules serves to increase their arrestin-binding capacity to ensure virtually complete arrestin translocation to microtubule-rich compartments (26) of the rod photoreceptor in the dark (6,19,21). Thus, arrestin tetramer not only serves as a “storage” form, but it also directly binds microtubules to facilitate the removal of arrestin away from the outer segment in dark-adapted photoreceptors. This ensures maximum gain of photoresponse by preventing premature signal termination by the extremely abundant arrestin. It is highly likely that this mechanism enables dark-adapted rods to generate a response to a single photon, while keeping enough stored arrestin to bind virtually every rhodopsin molecule in bright light.

3.5 How Rhodopsin Phosphorylation Level Regulates Arrestin Binding It was established more than 20 years ago that for high-affinity binding, arrestin requires that rhodopsin be both activated and phosphorylated (51). However, the effect of increasing the level of rhodopsin phosphorylation on arrestin binding remains controversial. Every conceivable model has been proposed: that arrestin

7315.indb 68

11/19/07 8:33:22 PM

69


affinity gradually increases with the level of rhodopsin phosphorylation (52,53) or that tight arrestin binding is an all-or-nothing event requiring a certain number of receptor-attached phosphates. This number was reported to be one (54,55), two (38,48,56), or three (16,57,58), and in some cases conflicting reports came from the same group (54,57). The use of a sensitive direct binding assay (described in the following text) revealed that arrestin independently recognizes the activation and phosphorylation state of rhodopsin, but its binding to light-activated phosphorhodopsin (P-Rh*) is at least 10 times higher than to dark P-Rh or unphosphorylated Rh* (figure 3.3). This remarkable arrestin selectivity for P-Rh* is explained by the current model of sequential multisite interaction (reviewed in references 59 to 61). This model, proposed in 1993 (38), is based on the hypothesis that arrestin has two primary binding sites: an activation sensor, which interacts with parts of rhodopsin that change conformation

Bound Arrestin, fmol

Step 1

Dark p-Rh P-Rh* Dark Rh Rh*

40

Step 2

Step 3

20

0

(a)

(b)

Figure 3.3 Exquisite arrestin selectivity for P-Rh* (a) is explained by the model of sequential multisite arrestin–rhodopsin interaction (b). Weak binding to any form of rhodopsin enables arrestin to “probe” its functional state. When arrestin encounters phosphorylated or activated rhodopsin, it binds either to the phosphorylated C-terminus or to the elements that change conformation upon activation via the arrestin phosphorylation-recognition or activation-recognition site, respectively (Step 1). If rhodopsin is only activated or phosphorylated, arrestin dissociates rapidly. When rhodopsin is phosphorylated and activated, both arrestin sites are engaged (Step 2). The simultaneous binding of the phosphorylation- and activationrecognition sites (termed primary binding sites) to respective parts of rhodopsin triggers a substantial conformational change in arrestin, which mobilizes additional potent binding sites for the interaction (Step 3). This conformation is termed the high-affinity receptor-binding state, in contrast to its basal low-affinity state. This complex multistep mechanism, which is shared by all arrestin proteins, makes arrestin binding to P-Rh* semiirreversible, so that arrestin dissociates only after Meta II decays to opsin, providing for high fidelity of signal termination. The binding in panel A was determined as described in Protocol 7 using 0.3 µg of the indicated form of rhodopsin and 50 fmol of radiolabeled arrestin. In panel B rhodopsin activation is indicated by lighter shading, and phosphorylation is indicated by three spheres on the rhodopsin C-terminus. Global conformational change in arrestin upon its binding to P-Rh* is also shown (Step 3).

7315.indb 69

11/19/07 8:33:23 PM

70


upon light activation, and a phosphate sensor, binding to rhodopsin-attached phosphates. Arrestin works as a coincidence detector, i.e., when both primary sites are simultaneously engaged (which can happen only when arrestin encounters P-Rh*), arrestin undergoes a transition into a high-affinity receptor-binding state involving a global conformational change (62) that results in the mobilization of additional binding sites (38). This model has been validated by the discovery of the main (63) and auxiliary (56) phosphate sensors and a convincing description of the mechanism of their function in the context of the arrestin crystal structure (33,64). Thus, in the context of this model the question can be restated, as follows: How many rhodopsinattached phosphates are necessary to activate the phosphate sensor in arrestin? A surprisingly large number of the exposed positively charged residues in rod arrestin have been shown to bind rhodopsin-attached phosphates (39,50,56,63,65), so that this issue cannot be unambiguously settled based solely on structural considerations. Obviously, direct comparison of arrestin binding to preparations of rhodopsin with different precisely determined phosphorylation levels is necessary to resolve this controversy. To this end, we used previously described procedures to phosphorylate rhodopsin in bovine rod outer segments by endogenous rhodopsin kinase (66), to purify differentially phosphorylated rhodopsin species, and to reconstitute them into liposomes (67,68). The actual phosphorylation level of rhodopsin in each fraction was carefully quantified by mass spectrometry, as described earlier (69,70). We ascertained that there is no appreciable difference in the orientation of reconstituted differentially phosphorylated rhodopsin by comparing the binding of all fractions to 4D2 antibody that recognizes the rhodopsin N-terminus (71) (which binds rhodopsin in the inside-out orientation). Finally, we used our extremely sensitive direct binding assay with radiolabeled arrestin to compare its interaction with rhodopsin carrying from 0 to 7 phosphates. The data reveal an interesting multithreshold mechanism: one rhodopsin-attached phosphate does not increase arrestin binding above the level observed with unphosphorylated Rh*. Two phosphates are the minimum number sufficient to facilitate arrestin interaction, three apparently activate arrestin fully, although four phosphates slightly but statistically significantly increase the binding, whereas additional phosphates do not appreciably enhance the binding further (figure 3.4) (70). These data indicate that in the context of the rod photoreceptor cell, which expresses only 2–3 rhodopsin kinase molecules (72,73) and up to 800 arrestin molecules (19,20) per 1000 rhodopsins, arrestin is likely to bind light-activated rhodopsin as soon as two to three phosphates are attached. Indeed, in vivo studies revealed the presence of rhodopsin species carrying one, two, and three phosphates, with negligible amounts of higher levels of phosphorylation (69). It has also been shown in transgenic mice expressing mutant rhodopsin with different numbers of phosphorylation sites that three are required to observe arrestin-dependent acceleration of photoresponse shutoff (16). Collectively, these data suggest that rhodopsin signaling is turned off in just 3–5 steps (2–4 phosphorylation events followed by arrestin binding), favoring the models that explain limited variability of the single-photon response by a small number of steps of rhodopsin inactivation (e.g., reference 74). However, the C-terminus of mammalian rhodopsin carries 6–7 phosphorylation sites, suggesting that their abundance is biologically important. Moreover, the elimination of even

7315.indb 70

11/19/07 8:33:23 PM

71

Arrestin Bound, fmol

How Rod Arrestin Achieved Perfection d

40

c

30

b

20 10

0 Phosphates 0

a 0

1

1

2

2

3

3

3

4

4

5

5

5

6

6

7

7

Figure 3.4 The effect of rhodopsin phosphorylation level on arrestin binding. Purified reconstituted phosphorhodopsin (0.15 µg) with the indicated level of phosphorylation was incubated with 50 fmol of radiolabeled arrestin in 50 µL for 5 min at 37°C in the light, and then bound arrestin was separated from free and quantified (Protocol 7). Binding data were analyzed using one-way ANOVA with phosphorylation level as a main factor, followed by a Bonferroni/Dunn post hoc test with correction for multiple comparisons. The binding levels were grouped (“a” through “d”) so that they were not significantly different within the group and significantly different from other groups on the same panel (p < 0.0001 in all cases).

one or two out of six sites in mouse rhodopsin clearly affects photoresponse kinetics, and the elimination of three substantially slows down rhodopsin inactivation (16). It is tempting to speculate that the large number of sites is necessary to increase the probability that transient relatively low-affinity binding of rhodopsin kinase to lightactivated rhodopsin is productive, i.e., results in the phosphorylation of at least one of the sites. This would become increasingly important with the progression of phosphorylation, e.g., after two sites are already phosphorylated, wild-type rhodopsin would retain four accessible targets, whereas the mutant carrying three sites would have only one left. In vivo comparison of the phosphorylation kinetics of wild-type and mutant rhodopsin is necessary to test this hypothesis. Protocol 7: Direct Binding Assay with Radiolabeled Arrestin This assay takes advantage of the expression of arrestins with high specific activity in cell-free translation (described in detail in references 35 and 75), which provides sufficient sensitivity to study high-affinity arrestin binding to P-Rh*, as well as low-affinity interactions with other forms of rhodopsin. As an added bonus, in vitro translated visual arrestin can be used without further purification, because it is the only radiolabeled protein in the translation mix. This relatively high-throughput assay is perfectly suited for the functional characterization of multiple mutant forms of arrestin (38,48,63,65), as well as for the direct comparison of wild-type arrestin binding to multiple rhodopsin preparations (figures 3.2 and 3.4) (70).

7315.indb 71

1. Store translated radiolabeled arrestin at –80°C, where it is stable for weeks. It survives several freeze–thaw cycles very well, much better than prolonged incubation at 4°C, so that it is advisable to thaw the sample “just in time” and then refreeze it immediately after taking the aliquot for the binding experiment.

11/19/07 8:33:24 PM

72


2. Dilute arrestin (50–100 fmol per assay, for 1–2 nM final concentration) with 50 mM Tris-HCl, pH 7.5, 50 mM potassium acetate, 0.5 mM MgCl2 (RB buffer) containing 0.5 mM DTT to 2–4 fmol/µL, and add 25 µL of this solution per tube. 3. Under dim red light, dilute Rh and P-Rh with the same buffer to 12 µg/mL and distribute between tubes (25 µL/assay, i.e., 0.3 µg or ~7.5 pmol per tube). 4. Incubate arrestin with the appropriate form of rhodopsin in the dark or with illumination at 37°C for 5 min. Precise timing is crucial, because of the rapid decay of light-activated rhodopsin. 5. Immediately cool the samples on ice. Separate bound and free arrestin by gel filtration at 4°C (under dim red light for dark rhodopsin) on a 2 mL Sepharose CL-2B column equilibrated with 10 mM Tris-HCl, pH 7.5, 100 mM NaCl (100/10). The membrane-containing fraction elutes in the void volume between 0.5 and 1.1 mL and is counted in a liquid scintillation counter. This is accomplished by having columns in a 10 × 10 rack that can be directly placed over a standard 10 × 10 box of scintillation vials. Samples are loaded on the columns and allowed to soak in. The column is then washed with 100 µL of 100/10 followed by 400 µL of 100/10 after the first portion enters the matrix, and then rhodopsin with bound arrestin is eluted into scintillation vials with 600 µL of 100/10. It is crucial that the columns for dark samples are run under dim red light. Determine nonspecific arrestin binding (using 0.3 µg of liposomes) and subtract. The columns for this assay can be reused many times, provided that after each experiment the columns are immediately washed with 4 × 3 mL of 100/10, and stored capped at 4°C in 100/10 buffer. If columns are not used for an extended period of time, the addition of 3 M NaCl to the storage buffer is advised to prevent microbial growth.

3.6 Stoichiometry of the Arrestin–Rhodopsin Interaction Based on the fact that rods faithfully respond to a single photon (73) and that a photon can only activate one rhodopsin (1), it was widely believed for a long time that arrestin interacts with an individual rhodopsin molecule (1,59,76). All arrestins in their basal state are elongated molecules consisting of two cuplike domains, each with a diameter of about 35 Å (32,33,39,41,42). Recent discoveries suggest that rhodopsin (77) and many other G-protein-coupled receptors (78) form dimers under certain circumstances. Dark (inactive) rhodopsin is the only receptor for which the crystal structure is known (79–81). Its relatively small cytoplasmic tip measures about 40 Å across. On the basis of this rather provocative geometry, an alternative model in which a single arrestin binds to two receptors in a dimer was proposed (82). The biological implications of these two models are profoundly different, so it is very important to resolve this issue experimentally. Luckily, rod photoreceptors provide a unique model in which the stoichiometry of the arrestin–receptor interaction can be determined in vivo because they express comparable amounts of rhodopsin and arrestin at very high levels unparalleled in any other cell type (5,19,83,84). As discussed earlier, light induces arrestin translocation

7315.indb 72

11/19/07 8:33:24 PM


73

from the inner to the outer segment, where it is retained by virtue of its binding to light-activated rhodopsin (6). It has been established that the expression of rhodopsin and arrestin can be genetically manipulated independently: hemizygous rhodopsin (Rh+/−) and arrestin (Arr+/−) mice express about half of these respective proteins as compared to wild-type animals (23,85). Thus, the extent of arrestin translocation in these genetic backgrounds can be used to determine the stoichiometry of the arrestin–rhodopsin interaction using exclusively wild-type proteins for this purpose. To this end, we bred Rh+/−, Arr+/−, and Rh+/−Arr+/− animals and compared the content of rhodopsin and arrestin in their retinas with that of wild-type mice (table 3.1). Both proteins were measured by quantitative Western blot in the homogenates of whole eyecups using the corresponding purified proteins to construct calibration curves. The results confirmed that the elimination of one allele reduces the expression by ~50%, so that the arrestin/rhodopsin ratio in wild-type and Rh+/−Arr+/− animals is similar (0.8:1 and 0.94:1, respectively), whereas in Arr+/− and Rh+/− mice, it is significantly shifted in the expected direction, to 0.38:1 and 1.74:1, respectively (table 3.1). If two rhodopsins were necessary to bind one arrestin, virtually complete arrestin translocation to the OS in bright light would only be expected in Arr+/− animals, but not in wild-type mice. In contrast, in the case of a one-to-one interaction, partial translocation would be expected only in Rh+/− animals that express more arrestin than rhodopsin. In agreement with previous reports (6,12–14,19,21,28), we invariably observed virtually complete arrestin translocation in the light-adapted retinas of wild-type mice, as well as other lines expressing more rhodopsin than arrestin (figure 3.5). Quantitative image analysis shows that in wild-type mice and the other two lines that express excess rhodopsin, 81–89% of arrestin translocates to the OS in the light (table 3.1) (20). In contrast, in Rh+/− mice, only about half of that amount of arrestin moves to the OS (table 3.1, figure 3.5). These data are consistent with one-to-one binding and cannot be reconciled with the model of rhodopsin dimer interacting with just one arrestin molecule. Based on the extent of arrestin translocation and the absolute expression levels of both proteins, one can calculate that in the light-adapted retinas of these mice, 65–83% of rhodopsin is occupied at steady state by bound arrestin (table 3.1), strongly supporting the one-to-one binding model. Notably, the rhodopsin occupancy in both Rh+/− lines (75–83%) exceeds the theoretical maximum for the one-to-two model (50%) even more than that in Rh+/+ animals. Interestingly, the reduction of rhodopsin concentration in the disks in Rh+/− mice has been shown to accelerate the kinetics of photoresponse, possibly by increasing the lateral diffusion of rhodopsin that facilitates its interactions with transducin (20). Higher rhodopsin occupancy by arrestin in Rh+/− lines than in the Rh+/+ lines suggests that the reduction of rhodopsin concentration in the disk membrane favors arrestin binding as well; i.e., the relief of rhodopsin “overcrowding” facilitates its interactions with all signaling proteins. Thus, arrestin translocation in every genetic background is only consistent with the one-to-one model of arrestin–rhodopsin binding in vivo. However, one can argue that in the photoreceptor cell other proteins could participate in arrestin retention in the OS, so that these in vivo data are not necessarily conclusive. To ascertain that other proteins present in live photoreceptors do not affect the translocation of arrestin to the OS, we reproduced similar arrestin/rhodopsin ratios (0.5, 1.1, 2.2, and

7315.indb 73

11/19/07 8:33:24 PM

7315.indb 74

0.40 ± 0.05 (7) 0.40 ± 0.03 (5) 0.18 ± 0.02 (4) 0.19 ± 0.01 (4)

Genotype

WT A+/– A+/–Rh+/– Rh+/–

0.32 ± 0.04 (7) 0.15 ± 0.02 (5) 0.17 ± 0.02 (4) 0.33 ± 0.02 (4)

Arrestin Content (nmol/retina) 0.80 0.38 0.94 1.74

Arr/RhMolar Ratio 1.8 ± 0.8 3.1 ± 1.1 1.4 ± 0.5 1.1 ± 0.4

Percentage Arr in the OS (dark) 81.4 ± 1.3 88.8 ± 0.5 89.1 ± 1.0 43.5 ± 6.5

Percentage Arr in the OS (light) 0.26 0.13 0.15 0.14

Arr in the OS (light) nmol

65 33 83 75

Rh Occupied by Arr (%)

Note: Mice with the indicated genotypes were dark-adapted overnight (dark) or exposed to 2700 lux for 1 h (light). One eyecup from each mouse was fixed and processed for immunohistochemistry (Protocol 3), whereas the other was homogenized for rhodopsin and arrestin quantification by Western blot (Protocol 8). The proportion of arrestin localized in the OS was quantified by the intensity of arrestin immunostaining in 10 images per animal from 3 to 5 animals per genotype per light condition. Means ± SD are shown. The data were analyzed by one-way ANOVA with genotype as a main factor. The rhodopsin content of A+/−Rh+/− and Rh+/− and the arrestin content of A+/− and A+/−Rh+/− were statistically different from all other genotypes (p < 0.0001) but were not different from each other. The amount of arrestin in the OS (nmol) was calculated by multiplying the arrestin content by the percentage of arrestin in the OS in the light. The percentage of rhodopsin occupied was determined by dividing this value by the total rhodopsin content.

Rh Content (nmol/retina)

Table 3.1 Light-Dependent Translocation of Arrestin to the Outer Segment as a Function of Arrestin and Rhodopsin Expression

74 Signal Transduction in the Retina

11/19/07 8:33:25 PM

75


3.3) in vitro with purified proteins carefully quantified by amino acid analysis. In these experiments, recombinant purified bovine arrestin was mixed in the dark with bovine rhodopsin in native disk membranes that was phosphorylated with endogenous rhodopsin kinase (66) and fully regenerated with 11-cis-retinal. The samples were illuminated for 5 min at 37°C, and then rhodopsin along with bound arrestin was pelleted by centrifugation. Samples containing the same amounts of arrestin and no rhodopsin were used as controls. Equal aliquots of the original samples, pellets, and supernatants were resolved by SDS-PAGE and stained with Coomassie blue, revealing clearly saturable arrestin binding (figure 3.6). To achieve high precision, OS IS ONL Genotype Arr:Rh

WT 0.8:1

Arr+/– 0.4:1

Arr+/–Rh+/– 0.9:1

Rh+/– 1.7:1

Figure 3.5 Complete arrestin translocation to the OS requires the presence of equivalent number of rhodopsin molecules. Mice with the indicated genotypes were dark-adapted overnight and exposed to 2700 lux for 1 h. Arrestin localization was visualized and quantified as described in Protocol 3. The positions of the outer (OSs) and inner segments (ISs) and of the outer nuclear layer (ONL) are indicated. Arrestin/rhodopsin expression ratios in each genotype are shown. Note that arrestin translocation is incomplete only when arrestin is expressed in excess over rhodopsin. 1.00

Pellet

3.3

2.2

1.1

Arr/Rh Ratio

0.5

Supernatant

Arrestin Bound

Load

0.75 0.50 0.25 0.00

(a)

0

2 1 Arrestin:rhodopsin Ratio

3

(b)

Figure 3.6 The stoichiometry of arrestin–rhodopsin interaction. (a) Purified bovine arrestin (142, 286, 572, and 858 pmol) was incubated for 5 min in the light with 10 µg (261 pmol) of phosphorhodopsin in 25 µL of 50 mM MOPS-Na, pH 7.2, 100 mM NaCl. Rhodopsin with bound arrestin was pelleted through a 50 µL cushion of the same buffer with 0.2 M sucrose. Equal aliquots of the sample before centrifugation (Load) (1/14), Pellet (1/4), and supernatant (Sup) (1/4) were subjected to SDS-PAGE. Arrestin was visualized by Coomassie blue staining. Arrestin/rhodopsin ratios are shown under the corresponding lanes. (b) The absolute amount of arrestin and rhodopsin in the pellet was measured by quantitative Western blot (Protocol 8). Arrestin binding is plotted as a function of the arrestin/rhodopsin molar ratio in the sample. Means + SD of two experiments are shown. The analysis of the binding data (using GraphPad Prizm) yields Bmax (saturation) at 0.99 ± 0.08 mol/mol.

7315.indb 75

11/19/07 8:33:26 PM

76


we measured the amounts of rhodopsin and arrestin in the original samples, pellets, and supernatants by quantitative Western blot with appropriate standards (Protocol 8). Rhodopsin was found to pellet quantitatively; the amount of arrestin pelleted in the absence of rhodopsin did not exceed 4.5%. Under these conditions, we observed clear saturation of rhodopsin by increasing concentrations of arrestin, reaching up to 0.9 mol of specifically bound arrestin per mol of rhodopsin (figure 3.6). These results are in perfect agreement with the in vivo data (figure 3.5). Because only arrestin monomer can bind rhodopsin (45), these data definitively establish that one arrestin molecule binds one rhodopsin molecule. Interestingly, even at a very high arrestin/rhodopsin expression ratio in Arr+/+Rh+/− mice, the proportion of arrestin-occupied rhodopsin “saturates” at ~80%, which almost exactly corresponds to the relative expression of these proteins in wild-type animals (table 3.1 and references 19 and 20). Apparently, mice express just enough arrestin to occupy all rhodopsin that is binding competent (i.e., light-activated, light-activated phosphorylated, dark phosphorylated, and phosphoopsin (38,70) (figure 3.2)) at equilibrium in fully light-adapted rods. Some of the rhodopsin in light-adapted OS exists as unphosphorylated opsin that does not bind arrestin and does not support arrestin translocation when the equilibrium is shifted toward this form by hydroxylamine treatment (6). This actually reduces the amount of rhodopsin competent to bind arrestin, further strengthening the conclusion that only the binding of arrestin by each individual rhodopsin can account for the observed extent of arrestin translocation in vivo. Protocol 8: Quantification of the Content of Individual Protein in Tissue Samples by Western Blot Sometimes it is important to actually measure the content of a particular protein in a tissue sample that can be expressed in absolute (weight or molar) units. This task becomes even more complicated when the absolute content of different proteins and their expression ratios must be established (20,86,87). This can be accomplished by Western blot only when a rigorous set of conditions is met. Here we describe these conditions, necessary reagents, and quality control procedures that ensure reliable quantitative results. Protein standards

7315.indb 76

1. First, one needs the pure proteins of interest, recombinant or purified from the tissue. The amounts of purified protein (>0.2 mg) must be sufficient to perform quality control experiments and have enough left for the actual work. The purity of each protein in preparations used to prepare the standards must be determined by SDS-PAGE followed by Coomassie blue staining. To achieve this, it is important to run different amounts of protein: “overloaded” lanes with 3–10 µg to detect even minor impurities, as well as lanes with 0.2–1 µg to ascertain that proteolytic products of a size similar to the full-length protein do not contaminate the sample. Purity below 95% is unacceptable. 2. Initial protein quantification by any standard method (e.g., Bradford), which is usually based on the comparison of the protein of interest with BSA,

11/19/07 8:33:26 PM


7315.indb 77

77

γ-globulin, or some other protein, must be checked by one of the “absolute” methods: sequence-calibrated UV absorbance or amino acid analysis. 3. For the sake of stability many proteins are kept in rather complex buffers with a variety of components, some of which absorb UV. Therefore, for spectroscopic measurements it is imperative that the spectrum of exactly the same buffer in which the protein is dissolved is also taken and subtracted. The best way to achieve this is to dialyze the protein extensively (>2 h for relatively small 0.1–0.5 mL samples) against the buffer, and then use the dialysis buffer in the beaker as a control. 4. Quantification by amino acid analysis requires the elimination of all buffer components from the sample, and in the case of membrane proteins (e.g., rhodopsin), the elimination of the lipid. The simplest way to accomplish this is to precipitate the protein by the addition of nine volumes of methanol and pellet it by centrifugation in tubes suitable for subsequent acid hydrolysis (e.g., 10 min at 13,000 rpm in 6 × 50 mm borosilicate tubes with appropriate holders in a standard tabletop microcentrifuge). The pellets should be washed with 90% methanol (0.2–0.5 mL) by vortexing and centrifugation, dried, and sent to a commercial or in-house protein chemistry facility. The proteins will be hydrolyzed with HCl (vapor phase, 150°C, 75 min) along with a standard amino acid mix (to control for the loss of amino acids under these conditions). Samples will then be dissolved in an appropriate buffer, derivatized, and aliquots of each sample will be analyzed by HPLC. The quantification of amino acids stable under the conditions of acid hydrolysis (Ala, Val, Leu, Phe, and Lys) should be used to calculate the amount of each protein in the sample based on its known amino acid composition. The numbers based on each of these amino acids must agree to within 5–7%. If they do not, the sample likely contains too many impurities to be used to prepare standards. 5. The standards should be prepared by serial dilution of the purified protein, which should be kept in aliquots at −80°C. In most cases the protein can be diluted in SDS sample buffer and stored frozen. Some proteins in SDS sample buffer, especially in very diluted solutions, “go bad” after a few freeze–thaw cycles, as manifested by a significant decrease in signal intensity, so that it becomes necessary to make fresh dilutions for each experiment. Therefore, it is advisable to compare fresh and refrozen standards of the protein of interest early on to test whether this problem exists. 6. The signal on the blot from the same amount of protein sometimes decreases in the presence of other proteins in the sample (possibly because of “shielding” of the antigen by proteins with the same electrophoretic mobility transferred to the membrane). Therefore, whenever possible, the standards should be mixed with the same amount of protein from the same tissue of corresponding knockout animals that are going to be present in the experimental samples. For example, retinal protein from arrestin and rhodopsin knockout mice was added to arrestin and rhodopsin standards, respectively, in the study described earlier (84). However, many proteins cannot be knocked out because in their absence mice are not viable, and for

11/19/07 8:33:26 PM

78


other species there are very few knockouts available. In this case the protein in each experimental sample should be run at 2–3 different concentrations to ascertain that the signal increases linearly with the load. Quantitative Western blot

7315.indb 78

1. After standard disk-electrophoresis, we transfer the proteins to Immobilon-P (Millipore) membrane and block with 5% nonfat dry milk in TBS containing 0.1% Tween-20 (TBST) for 30–60 min at 37°C or for 60 min at room temperature with gentle rocking. 2. Several pilot experiments are usually required to establish conditions under which the blot is actually quantitative. First, it is necessary to establish the linear range of the signal by running 7–10 standards containing from 10 pg to 10 ng of the protein. Usually, this has to be done more than once (adjusting the range as you go) with different dilutions of primary antibody to optimize this parameter at the same time. In many cases the signal is linear only in a fairly narrow range and only at certain antibody concentration, both of which may be very different depending on the antibody, (e.g., in our hands rhodopsin signal is linear between 0.1 and 1.5 ng/lane at 1:3,000 dilution of 4D2 antibody (71), whereas arrestin signal is linear between 10 and 100 pg/lane at 1:10,000 dilution of F4C1 antibody (40). In some cases incubation with the primary antibody overnight at 4°C works better than 1 h at room temperature, but in other cases the opposite is true. Changes in the dilution or incubation time of the secondary antibody cannot compensate for the nonlinearity of primary antibody binding. Therefore secondaries should be used at a standard concentration, (e.g., we use 1:12,000 dilution of HRP-conjugated antirabbit and antimouse antibodies from Jackson ImmunoResearch) and incubate for 1 h at room temperature. 3. The great majority of commercial antibodies marketed as “specific” on the strength of a blot of the material from cells overexpressing the protein in question label more than one band in real-life tissue samples. Sometimes, only the comparison with the standard allows the identification of the specific band among others. It is often worthwhile to adjust the acrylamide concentration in the gel to ensure optimal separation of the band of interest from the nearest nonspecific bands. This will greatly facilitate subsequent quantification and improve its precision. The pattern of nonspecific bands may be very different in different tissues or cell lines. 4. Working solutions of many (but not all) primary antibodies can be frozen at −20°C and reused several times. This is only possible if the antibody is diluted in TBST supplemented with 2% BSA. As a rule, freezing antibodies in TBST with nonfat dry milk kills them. If a working solution of the primary antibody is to be reused, the milk from the blocking solution should be thoroughly washed away with TBST before the addition of the antibody. It is advisable to test whether the antibody survives freezing by comparing the signal on two identical blots incubated with fresh and thawed antibody solution.

11/19/07 8:33:27 PM


79

5. After incubation with HRP-conjugated secondary antibody we develop the blot with SuperSignal West Pico reagent (Pierce) for 1–1.5 min, and then expose it using SuperRX x-ray film (Fujifilm). For quantification purposes, it is imperative to obtain exposures where the bands in the standards and samples are not saturated (“pretty” blots with thick black bands that look good in papers are not quantitative). Often, it is useful to get two or more different exposures in which the bands to be used for quantification are “gray.” 6. A series of standards should be run on every blot. For quantification we use VersaDoc with QuantityOne software (Bio-Rad). Background (signal in an equal “empty” area, which is never zero even after background subtraction) should be subtracted from the integrated signal in the band (volume = intensity × mm2). The specific signal from standards should be plotted as a function of protein amount. This calibration curve (or at least the part of it that will be used) should be linear. Bands in experimental samples with a signal that falls within the linear range of the calibration curve can be used for quantification. It is very important to quantify each protein in each sample on at least two independent blots. These numbers should agree within 10%.

3.7 Release of Bound Arrestin To ensure high fidelity of quenching, arrestin has to stay bound as long as rhodopsin is in a functional state that has activity toward transducin. Indeed, arrestin was found to be the most slowly released protein among all that bind rod outer segment membranes in a light-dependent manner (88). The half-life of the complex at 0°C exceeds 2 h (38,70), suggesting that even at physiological temperature it would last for minutes, i.e., long enough for Meta II decay. Ultimately, to return the system to the ground state, arrestin must be released, and rhodopsin has to be dephosphorylated and regenerated with 11-cis-retinal. Because arrestin precludes rhodopsin dephosphorylation (89), it apparently has to dissociate first. The model of sequential multisite arrestin binding (figure 3.3) suggests that rhodopsin deactivation (decay of Meta II accompanied by the corresponding conformational changes in opsin) decreases arrestin affinity, thereby facilitating its release. Although in live photoreceptors arrestin translocation back to the inner segment in the dark occurs at the same rate as rhodopsin dephosphorylation (6), this correlation does not tell which event is the cause and which is the effect. However, dramatic acceleration of arrestin release by the addition of hydroxylamine (that facilitates Meta II decay) to the eyecups (6) indicates that the release of retinal from Meta II promotes arrestin dissociation. Interestingly, the release of retinal and bound arrestin may be interdependent events (70,90), and this area certainly requires further investigation.

3.8 Putting the Pieces Together: The Functional Cycle of Arrestin in Rods To better understand what the affinities of arrestin for rhodopsin, microtubules, and itself actually tell us, we need to look at arrestin interactions with various partners

7315.indb 79

11/19/07 8:33:27 PM

80


from a thermodynamic point of view. The interaction of two proteins is a chemical reaction. The Gibbs standard free energy change, ∆G°, for a reaction at constant temperature and pressure is given by: ΔG° = ΔH° − TΔS° where ΔH° is the enthalpy difference between reactants (free proteins in this case) and products (protein–protein complex) in their standard states, i.e., at concentrations of 1 M and at a specified temperature (usually 25°C); ΔS° is the entropy difference between reactants and products (in their standard states); and T is the temperature (in degrees Kelvin). The association of two proteins, similar to the association of a ligand with a receptor, is a second-order reaction (A + B = AB). The reaction proceeds when the product, complex AB, has less free energy than the reactants (free proteins A and B), so that a negative value of ΔG° is thermodynamically favored. The enthalpy term ΔH° shows whether heat is being released (ΔH° < 0) or absorbed (ΔH° > 0) during the reaction. The changes in enthalpy usually reflect the formation of intermolecular bonds (ion pairs, hydrogen bonds, dipole–dipole interactions). The entropy term ΔS° is intuitively perceived as an expression of a degree of randomness of the system. Strictly speaking, it represents the number of equivalent energy states available to a particular molecular species, and the changes in entropy often reflect the participation of hydrophobic interactions (because the removal of hydrophobic side chains from water environment yields favorable changes of entropy). For a second-order reaction, the free energy for association, ΔG°, is related to the equilibrium association constant as: ΔG° = −RT lnK A, where R is the gas constant (1.99 cal/mol × degree); T is the temperature (in degrees Kelvin); and K A is the equilibrium association constant (M−1); K A = 1/K D. Because a decrease in ΔG° drives the reaction forward, high-affinity interactions (large K A, i.e., small K D) are always favored over lower-affinity interactions. Thus, arrestin binding to rhodopsin (K D < 1 nM) is much more favorable than its binding to microtubules (K D > 50 µM) or self-association (dimerization [K D1] and tetramerization [K D2] constants 37.2 and 7.4 µM, respectively). Thus, arrestin will self-associate and bind MTs only when more “attractive” partners (P-Rh*, Rh*, dark P-Rh, P-opsin; cf. figure 3.2) are absent, i.e., in the dark. By the same token, self-association of arrestin monomers is somewhat more favorable than microtubule binding, so that arrestin is more likely to tetramerize first and bind MTs later. The ability of the tetramer to bind MTs without dissociating (45) greatly favors this scenario. The kinetics of arrestin binding to itself and other partners and the rate of its dissociation from these complexes are of utmost biological importance. This information can also be derived from the affinity of these interactions, because K D = k–1/k1, where

7315.indb 80

11/19/07 8:33:27 PM


81

K D is the equilibrium dissociation constant, and k1 and k–1 are kinetic association and dissociation constants. The association rate constant is proportional to the rate of diffusion, and because both rhodopsin and microtubules do not move between the compartments of the photoreceptor cell (at least on the subsecond time scale of photoresponse), we only need to take into account the rate of arrestin diffusion. The diffusion-limited association rate constant for ~45 kDa arrestin monomer can be estimated as 1 × 106 M−1sec−1 (91). The estimates of the amount of arrestin present in the outer segment in the dark vary from 1 to 3% (6,20) to 95% pure Pγ or mutant per liter of culture.

Note

1. This work was supported by National Institutes of Health Grant EY-10843.

References

7315.indb 96

1. Beavo, J.A. Cyclic nucleotide phosphodiesterases: functional implications of multiple isoforms, Physiol Rev, 75, 725, 1995. 2. Francis, S.H., Turko, I.V., and Corbin, J.D., Cyclic nucleotide phosphodiesterases: relating structure and function, Prog Nucl Acid Res Mol Biol, 65, 1, 2001. 3. Arshavsky, V.Y., Lamb, T.D., and Pugh, Jr., E.N., G proteins and phototransduction, Annu Rev Physiol, 264, 153, 2002. 4. Chabre, M. and Deterre, P., Molecular mechanism of visual transduction, Eur J Biochem, 179, 255, 1989. 5. Muradov, K.G., Granovsky, A.E., Schey, K.L., and Artemyev, N.O., Direct interaction of the inhibitory γ-subunit of rod cGMP phosphodiesterase (PDE6) with the PDE6 GAF-A domains, Biochemistry, 41, 3884, 2002. 6. Thompson, J. and Appleman, M.M., Multiple cyclic nucleotide phosphodisterase activities from rat brain, Biochemistry, 10, 311, 1971. 7. Hurley, J.B. and Stryer, L., Purification and characterization of the gamma regulatory subunit of the cyclic GMP phosphodiesterase from retinal rod outer segments, J Biol Chem, 257, 11094, 1982. 8. Yarfitz, S. and Hurley, J.B., Transduction mechanisms of vertebrate and invertebrate photoreceptors, J Biol Chem, 269, 14329, 1994. 9. Qin, N., Pittler, S.J., and Baehr, W., In vitro isoprenylation and membrane association of mouse rod photoreceptor cGMP phosphodiesterase α and β subunits expressed in bacteria, J Biol Chem, 267, 8458, 1992. 10. Qin, N. and Baehr, W., Expression and mutagenesis of mouse rod photoreceptor cGMP phosphodiesterase, J Biol Chem, 269, 3265, 1994.

11/19/07 8:34:52 PM

Function and Regulation of PDE6

97

11. Piriev, N.I., Yamashita, C., Samuel, G., and Farber, D.B., Rod photoreceptor cGMPphosphodiesterase: analysis of α and β subunits expressed in human kidney cells, Proc Natl Acad Sci USA., 90, 9340, 1993. 12. Sung, B.J., Hwang, K.Y., Jeon, Y.H., Lee, J.I., Heo, Y.S., Kim, J.H., Moon, J., Yoon, J.M., Hyun, Y.L., Kim, E., Eum, S.J., Park, S.Y., Lee, J.O., Lee, T.G., Ro, S., and Cho, J.M., Structure of the catalytic domain of human phosphodiesterase 5 with bound drug molecules, Nature, 425, 98, 2003. 13. Huai, Q., Liu, Y., Francis, S.H., Corbin, J.D., and Ke, H. Crystal structures of phosphodiesterases 4 and 5 in complex with inhibitor 3-isobutyl-1-methylxanthine suggest a conformation determinant of inhibitor selectivity, J Biol Chem, 279, 13095, 2004. 14. Granovsky, A.E., Natochin, M., McEntaffer, R.L., Haik, T.L. Francis, S.H., Corbin, J.D., and Artemyev, N.O., Probing domain functions of chimeric PDE6α/PDE5 cGMPphosphodiesterase, J Biol Chem, 279, 24485, 1998. 15. Muradov, K.G., Boyd, K.K., Martinez, S.E, Beavo, J.A., and Artemyev, N.O. The GAF-A domains of rod cGMP-phosphodiesterase 6 determine the selectivity of the enzyme dimerization, J Biol Chem, 278, 10594, 2003. 16. Muradov, K.G., Boyd, K.K., and Artemyev N.O., Analysis of PDE6 function using chimeric PDE5/6 catalytic domains, Vision Res, 46, 860, 2006. 17. Ovchinnikov, Yu. A., Lipkin, V.M., Kumarev, V.P., Gubanov, V.V., Khramtsov N.V., Akhmedov, N.B., Zagranichny, V.E., and Muradov, K.G., Cyclic GMP phosphodiesterase from cattle retina. Amino acid sequence of the γ-subunit and nucleotide sequence of the corresponding cDNA, FEBS Lett, 204, 288, 1986. 18. Brown, R.L. and Stryer, L., Expression in bacteria of functional inhibitory subunit of retinal rod cGMP phosphodiesterase, Proc Natl Acad Sci USA., 86, 4922, 1989. 19. Skiba, N.P., Artemyev, N.O., and Hamm, H.E., The carboxyl terminus of the γ-subunit of rod cGMP phosphodiesterase contains distinct sites of interaction with the enzyme catalytic subunits and the α-subunit of transducin, J Biol Chem, 278, 10594, 1995. 20. Slepak, V.Z., Artemyev, N.O., Zhu,Y., Dumke, C.L., Sabacan, L., Sondek, J., Hamm, H.E., Bownds, M.D., and Arshavsky, V.Y., An effector site that stimulates G-protein GTPase in photoreceptors, J Biol Chem, 278, 14319, 1995. 21. Artemyev, N.O., Natochin, M., Busman, M., Schey, K.L., and Hamm, H.E., Mechanism of photoreceptor cGMP phosphodiesterase inhibition by its γ-subunits, Proc Natl Acad Sci USA., 93, 5407, 1996.

7315.indb 97

11/19/07 8:34:53 PM

7315.indb 98

11/19/07 8:34:53 PM

5

Biochemical Characterization of Phototransduction RGS9-1–GAP Complex Qiong Wang and Theodore G. Wensel

Contents 5.1 5.2

Introduction.................................................................................................. 100 Expression and Purification of Recombinant Components, Fragments, and Mutants of the RGS9-1 Complex.......................................................... 100 5.2.1 Expression of RGS9-1 Fragments in E. coli..................................... 101 5.2.2 R9AP Expression and Purification in E. coli.................................... 102 5.2.3 Expression of the RGS9-1 Complex with Gβ5 in Insect Cells Using Baculovirus............................................................................. 103 5.3 Reconstitution of Full-Length R9AP into Lipid Vesicles............................ 105 5.3.1 Reconstitution of Purified R9AP into Vesicles Containing Phospholipids Only or Phospholipids and Rhodopsin...................... 105 5.4 Single-Turnover GAP Assay for Purified Recombinant or Retinal Proteins........................................................................................................ 106 5.5 Quantification of ROS Proteins Involved in GTPase Regulation by Quantitative Immunoblotting....................................................................... 108 5.6 Immunoprecipitation of the GAP Complex................................................. 111 5.7 Localization of Proteins in Rod Outer Segments by Subcellular Fractionation and Immunofluorescence....................................................... 114 5.7.1 Gradient Copurification: A General Way to Determine Whether a Protein Is Localized to ROS or a Contaminant from Other Parts of Retina.................................................................................. 114 5.7.2 Protein Localization by Immunofluorescence.................................. 116 5.8 Phosphorylation of RGS9-1.......................................................................... 117 References............................................................................................................... 119

99

7315.indb 99

11/19/07 8:37:13 PM

100


5.1 Introduction A major player in the regulation of timing and sensitivity of photoresponses is the GTPase accelerating protein, RGS9-1 (1–4). RGS9-1 binds to the α subunits of the rod and cone G-proteins, Gαt1 and Gαt2, when they are in their activated GTP-bound conformations, and speeds up the rates at which they hydrolyze GTP, thereupon returning to the inactive GDP-bound conformations. This conceptually simple function is complex in its biochemical details and regulation. In addition to RGS9-1, two other subunits, Gβ5L (5–9) and R9AP (RGS9-1 anchor protein) (10–12), are required for the function and stability of the GTPase accelerating protein (GAP) complex, and the inhibitory subunit of the photoreceptor effector enzyme, PDE6γ, dramatically enhances the affinity of this complex for Gαt-GTP. The activity of the complex and the time resolution of vision depend on its concentration in the cells (13), and the concentrations are very different in rods and cones (1,14). The Gβ5 gene is subject to alternative splicing, and although only one splice variant, G β5L, is found in rods (15), both Gβ5L and Gβ5S are found associated with RGS9-1 in cones (14). RGS9-1 is also subject to Ca2+-regulated phosphorylation by protein kinase C (16,17), and may be further regulated by phosphoinositides. Levels of this complex have been shown to be essential for the timely recovery of photoresponse by loss-of-function studies in mice (18–21) and humans (22), and those proteins also control the rate-limiting step in vision confirmed by a recent discovery that overexpression of the whole complex speeds up the response (13). This chapter will focus on biochemical techniques used to characterize the molecular mechanisms of RGS9-1–GAP complex in phototransduction kinetics and their regulation.

5.2 Expression and Purification of Recombinant Components, Fragments, and Mutants of the RGS9-1 Complex Characterization of the RGS9-1–Gβ5–R9AP complex requires expression and purification of recombinant proteins, with either wild-type or reengineered sequences. Three major heterologous expression systems have been previously tested, and each has its own advantages. The choice of expression system depends on the requirement of these proteins for different applications: in vitro biochemical assay, generation of antibodies, or functional studies of domains and specific sites. There are some special considerations that must be taken into account when working with these proteins. One is that any RGS9-1 construct containing the GGL domain must be coexpressed with Gβ5, if soluble and functional protein is to be obtained, and, conversely, no success has been achieved in expression of functional G β5 without coexpression of a GGL-domain-containing construct. R9AP is a transmembrane protein, in contrast to the RGS9-1–Gβ5 complex, so conditions for its expression and purification are necessarily different. Although R9AP is required for stability of both RGS9-1 and the long isoform of Gβ5 in vivo in photoreceptors, in vitro expression systems do not require coexpression. All three proteins have been successfully expressed in both insect cells and mammalian tissue culture cells, and they appear to form functional complexes. The RGS9-1–Gβ5 complex can be expressed and purified without R9AP,

7315.indb 100

11/19/07 8:37:14 PM

Biochemical Characterization of Phototransduction RGS9-1–GAP Complex

101

and then recombined either with full-length R9AP reconstituted into lipid vesicles, or with the cytoplasmic domain of R9AP, which binds tightly to RGS9-1 without the need for lipid or detergent.

5.2.1 Expression of RGS9-1 Fragments in E. coli The following procedure describes the expression and purification of either fulllength RGS9-1 or fragments for antibody generation and functional assays (2,23,24). Both His6 tags (pET vectors, Novagen) and glutathione-S-transferase (GST, pGEX vectors, GE) tags have been used successfully. Neither full-length RGS9-1 nor the RGS domain-containing fragments (i.e., containing residues 291–418) of RGS9-1 are initially soluble when overexpressed in E. coli. They are found primarily in inclusion bodies, so simply lysing the cells and washing the inclusion bodies gives a good first step of purification. Subsequently the proteins can be extracted under denaturing conditions. His6-tagged and GST-tagged proteins both renature well into active form upon dialysis to remove denaturants. The His-tagged protein is purified by metal ion affinity chromatography under denaturing conditions, and then renatured by step dialysis, whereas the GST-tagged protein is renatured by dialysis prior to affinity purification using glutathione beads. In contrast, full-length RGS9-1 does not fold properly without Gβ5, so for this protein, E. coli expression is useful only to produce denatured protein for antibody production or for use as a standard in gel-based (Coomassie-blue staining or Western blotting) quantification procedures. Protocol 1: Bacterial Expression and Purification of Tagged RGS9-1 Fragments

7315.indb 101

1. Cells (BL21 (DE3) pLysS) are freshly transformed with plasmid using standard procedures, and incubated overnight at 37°C on LB-ampicillin/ chloramphenicol plates. The next day multiple colonies are collected and used to inoculate the LB medium. A typical volume is 1 L. pLysS strains, expressing T7 lysozyme, are useful for efficient disruption of cell walls. However, when a Microfluidizer® (Microfluidizer Corp.) is used to break open the cells, BL21 (DE3) without pLysS can be used, and no chloramphenicol is needed. 2. For vectors with the T7-lac promoter (e.g., pET-14B-derived expression plasmids), 1 mM IPTG (isopropyl-beta-d-thiogalactopyranoside) is used to induce expression when the cells have reached an OD600 of 0.6, and growth is allowed to continue at 37°C if the intent is to extract protein from inclusion bodies, or at 25°C or even lower temperatures if the intent is to optimize the amount of initially soluble protein. For vectors with the T7 promoter (e.g., pGEX-2TK), 0.1 mM IPTG is used to induce expression. In general, it is a good idea to remove samples for SDS-PAGE assessment of protein expression at different time points. 3. Cells are harvested by 4000 × g centrifugation (20 min, 4°C), and pellets are sonicated and washed twice with lysis buffer: 50 mM sodium phosphate, pH 7.4, 300 mM NaCl, 1 mM dithiothreitol (DTT). The supernatants from these washes can then be used for affinity purification on immobilized

11/19/07 8:37:14 PM

102


metal or glutathione beads, following the manufacturer’s instructions. However, for most RGS9-1 fragments, a much higher yield of soluble protein is obtained by extracting and renaturing proteins from inclusion bodies. 4. The inclusion bodies (pellets from previous washes) are solubilized in 35 mL/L culture of guanidinium HCl buffer: 6 M guanidinium HCl, 100 mM sodium phosphate, 10 mM Tris, pH 8.0. The samples are rotated at room temperature for 2 h. 5. For His-tagged proteins, the solubilized protein is then applied to nickel nitrilotriacetic (Ni2+–NTA) beads (Qiagen) and purified under denaturing conditions according to the manufacturer’s instructions. The purified protein is then diluted to a concentration of 0.1 mg/mL with urea buffer: 8 M urea, 100 mM sodium phosphate, 10 mM Tris, pH 8.0. The protein is then dialyzed overnight at 4°C against a 10,000-fold larger volume of renaturing buffer: 50 mM sodium phosphate, pH 7.4, 300 mM NaCl, 10% glycerol, 0.1% 2-mercaptoethanol. Insoluble protein is removed by centrifugation at 20,000 × g (20 min 4°C), and the protein in the supernatant is concentrated by centrifugal ultrafiltration (Centricon, Millipore). Most RGS9-1 fragments show a tendency to aggregate at concentrations above 10 µM, so there is not much point in trying to exceed this concentration. 6. For GST-tagged proteins, the crude solubilized inclusion bodies are renatured as described earlier to renature the GST moiety, prior to their being loaded onto glutathione beads and purified according to the manufacturer’s instructions. 7. Full-length RGS9-1 does not renature, but instead forms precipitates when denaturants are dialyzed away. The denatured protein can be stored at −20°C after adding glycerol to 40% (v/v) and used as a standard for quantifying recombinant proteins expressed by other means using Coomassie staining on gels or quantitative immunoblotting.

5.2.2 R9AP Expression and Purification in E. coli R9AP is a transmembrane protein, and thus the full-length form, containing its single C-terminal transmembrane helix, is expressed in insoluble form in bacteria. Fortunately, however, it is one of the few eukaryotic membrane proteins that can be solubilized from bacterial membrane pellets using nondenaturing detergents, purified, and reconstituted in functional form into lipid vesicles. Moreover, the cytoplasmic domain of R9AP can be expressed as a soluble fragment in bacteria. In both cases, the protein has been expressed in pET14b and purified using N-terminal His tags. Immobilized metal-ion affinity chromatography is not sufficient to obtain pure protein, so an additional step of ion exchange chromatography is used. Protocol 2: Expression of R9AP and its Soluble Fragments in E. coli and Purification by Affinity and Ion Exchange Chromatography

7315.indb 102

1. Transformation of BL21 (DE3) pLysS cells is carried out as described earlier for RGS9-1 fragments. Cells are induced when OD600 = 0.6–0.8 with 0.3 mM IPTG, and allowed to grow for 4–5 h at 30°C.

11/19/07 8:37:14 PM


103

2. Cells are harvested by centrifugation at 4000 × g for 20 min at 4°C, and the pellets sonicated with lysis buffer (25 mM Tris, pH 8.0, 300 mM NaCl, 20 mM imidazole, 2 mM DTT, and ~20 mg/L PMSF, phenylmethylsulfonyl fluoride). 3. His-tagged bovine R9AP fragments lacking the C-terminal transmembrane helix (His-R9AP-∆C, amino acids 1–212 or His-R9AP-∆C2, amino acids 1–191) are largely soluble because they do not contain the C-terminal transmembrane helix. In these cases, cell debris and insoluble materials are removed by centrifugation at 24,000 × g for 30 min, and the soluble fragments in the supernatant are purified under native conditions with immobilized metal ion beads following the manufacturer’s protocol, and eluted with 200 mM imidazole in the lysis buffer. The purest fractions are combined and dialyzed against dialysis buffer (10 mM Tris-HCl, 1 mM DTT, pH 8.0). The protein is not exceptionally pure at this point, so it is applied onto a strong anion exchange column (e.g., POROS HQ HPLC anion exchange column, Applied Biosystems), washed with 20 mM NaCl in the dialysis buffer, and eluted with a linear gradient from 20 to 200 mM NaCl in the same buffer (at a flow rate of 2–3 mL/min for the POROS HPLC column). About 10 mg of protein can be obtained from 1 L of culture. 4. His-tagged mouse full-length R9AP is insoluble and is found in pellets after centrifugation of the cell lysates as described earlier. The proteins are extracted with 4% sodium cholate in lysis buffer for 0.5–1 h at 4°C with gentle agitation, followed by centrifugation; this procedure is repeated 3–4 times, yielding >70% of total His-mR9AP extracted in a soluble form. The pooled detergent-solubilized R9AP is purified using Ni2+-NTA beads in 4% sodium cholate in the lysis buffer according to the manufacturer’s protocol. Routinely, at least 5 mg R9AP with 90–95% purity can be obtained from 2 L E. coli cultures. Reconstitution procedures are described in section 5.3.

5.2.3 Expression of the RGS9-1 Complex with Gβ5 in Insect Cells Using Baculovirus The most useful expression system so far for production of functional full-length RGS9-1 and its partner Gβ5 is baculovirus-directed expression in insect cells. Attempts to express each separately have produced low yields of most insoluble proteins, for which no activity has been detected; however, coexpression using simultaneous infection with two viruses produces useful amounts of soluble, active protein complex. It is generally necessary to test different ratios of RGS9-1 virus to Gβ5 virus to determine the one that produces the highest yield of the complex; this ratio will vary over time if the titers of the virus stocks decline at different rates, so it is advisable to retiter the virus from time to time. The following procedure (protocol 3) uses adherent cells growing in plates. Alternatively, cells can be grown in suspension as described (25). There are two known splice variants of Gβ5, Gβ5L and Gβ5S. In insect cells, viruses encoding the long-variant Gβ5L produce a mixture of Gβ5L and Gβ5S, with the cells apparently using the first methionine residue of Gβ5S as an alternative translation start site.

7315.indb 103

11/19/07 8:37:15 PM

104


Protocol 3: Insect Cell Expression and Purification of RGS9-1/Gβ5

1. If it is necessary to produce new virus, recombinant baculoviruses are isolated after cotransfection of the linearized BaculoGold viral DNA (PharMingen) and the recombinant transfer vector pVL1392 (PharMingen) with the proper insert into Sf9 cells following the manufacturer’s protocol. 2. The cells are grown as monolayers in 150 mm culture dishes in InsectXpress medium (Bio Whittaker) supplemented with 8% fetal bovine serum and 10 µg/mL gentamycin. They are coinfected with two types of recombinant viruses containing RGS9-1 or G β5L (the MOI [multiplicity of infection]) for both viruses is approximately 1:1) at 80% confluency and the cells are harvested 48 h later. 3. Cells are removed from the plates, collected by centrifugation, and the cell pellet is resuspended in ice-cold lysis buffer (50 mM Tris-HCl, pH 8.0, 500 mM NaCl, 1 mM DTT, and 1% Nonidet P-40) with freshly added protease inhibitors (0.03 mg/mL leupeptin, 0.017 mg/mL pepstatin A, 0.005 mg/ mL aprotinin, 0.03 mg/mL lima bean trypsin inhibitor, and ~20 mg/L solid PMSF) and mixed by rocking for 1 h at 4°C. The cell suspensions are sonicated on ice, centrifuged at 20,000 × g for 30 min at 4°C, and the supernatants are collected. 4. For His-tagged proteins, the supernatants are supplemented with 20 mM imidazole and loaded onto a Ni2+-NTA agarose column, washed with lysis buffer, and then with GAPN-H buffer (10 mM HEPES, pH 7.4, 100 mM NaCl, 2 mM MgCl2, 1 mM DTT, and ~20 mg/L PMSF) containing 20 mM imidazole, and then eluted with 250 mM imidazole in GAPN-H buffer. 5. For GST-tagged proteins, the supernatants are loaded to a glutathione-sepharose 4B column, washed with lysis buffer followed by GAPN-H buffer (10 mM HEPES, pH 7.0, 100 mM NaCl, 2 mM MgCl2, 1 mM DTT, and ~20 mg/L PMSF) and eluted with 40 mM glutathione in GAPN-H buffer. 6. Untagged Gβ5 expressed in the cells consistently copurify with RGS9-1 after affinity chromatography (figure 5.1). kDa 98 67 43 35 29 25

Fract:1

2

3

4

5

6

7

8

9 GST-RGS9 Gβ5L Gβ5S GST

19 13

Figure 5.1 Coomassie-blue-stained SDS-PAGE gel of glutathione elution fractions from GST-RGS9-1-G β5 expressed in insect cells. After the column had been thoroughly washed with buffer, as described in protocol 3, glutathione was added to the buffer, and fractions collected, beginning with Fraction 1.

7315.indb 104

11/19/07 8:37:15 PM


105

5.3 Reconstitution of Full-Length R9AP into Lipid Vesicles R9AP appears to have several roles in regulating RGS9-1 function. These include ensuring the stability of the RGS9-1–Gβ5L complex (20), controlling its localization to the outer segments (26), anchoring it to the disk membrane (10,11) where it interacts with Gαt-GTP and PDE6, and enhancing the catalytic activity of RGS9-1 (27). To study its membrane-dependent functions using purified proteins, it is necessary to reconstitute it into lipid vesicles. This is conveniently carried out with protein purified from E. coli in nondenaturing detergents, as described earlier, and well-defined phospholipid mixtures. This procedure can be carried out with or without the simultaneous presence of detergent-purified rhodopsin, and remarkably homogeneous vesicles as determined by cryoelectron microscopy are obtained (11).

5.3.1 Reconstitution of Purified R9AP into Vesicles Containing Phospholipids Only or Phospholipids and Rhodopsin Protocol 4: Rhodopsin Purification

1. Buffers used are lysis buffer (300 mM NaCl, 25 mM Tris, pH 8.0); GAPN-H buffer (100 mM NaCl, 2 mM MgCl2, 10 mM HEPES pH 7.4); ConA buffer (300 mM NaCl, 50 mM Tris-HCl, pH 7.0, 1 mM CaCl2, 1 mM MgCl2, 1 mM MnCl2); high-salt buffer (1 M NH4Cl, 10 mM HEPES, pH 7.4, 2 mM MgCl2). DTT, to 1 mM, and solid PMSF, to 20 mg/L, were added to each buffer just before use. 2. Purification of rhodopsin uses a modified version of a published procedure (28), with all procedures carried out either under dim red light or using nearinfrared illumination and infrared image-converting goggles. Sepharose beads containing immobilized concanavalin A (Con A) are stabilized before use to prevent bleeding off of Con A during elution by treatment with 0.05% glutaraldehyde in 250 mM NaHCO3 and prepared as described (28). Rod outer segments are prepared using a standard discontinuous sucrose gradient procedure (29), and extracted twice with high-salt buffer at a concentration of 15 µM rhodopsin or lower. They are then washed twice at the same concentration with Con A buffer, and the pellets are solubilized in Con A buffer containing 4% (w/v) sodium cholate to the same concentration. The supernatant is loaded onto the column, which is washed first with 10 column volumes of Con A buffer containing 4% sodium cholate, and then with 3 column volumes of the same buffer supplemented with 300 mM α-methyl mannoside. The eluant is concentrated by ultrafiltration to obtain a final rhodopsin concentration (determined by absorbance at 500 nM) of 2–3 mg/mL. Protocol 5: Reconstitution of R9AP with or without Rhodopsin in Lipid Vesicles

7315.indb 105

1. A phospholipid solution is prepared by first mixing chloroform solutions of lipids from Avanti Polar Lipids or Molecular Probes at a mass ratio of

11/19/07 8:37:16 PM

106


phosphatidylcholine:phosphatidylethanolamine:phosphatidylserine:rhodamine-labeled phosphatidylethanolamine = 50:35:1:0.43, drying it under a stream of argon, and then dissolving it in sufficient lysis buffer containing 4% sodium cholate to achieve a final lipid concentration of 20 mg/mL. The mixture is sonicated as necessary under argon and on ice to achieve a homogeneous solution. Then, cholate solutions of either rhodopsin, or Histagged R9AP, or both are added to the lipid solution to achieve a lipid-toprotein mass ratio between 20:1 and 40:1. 2. The same procedure without added proteins produces protein-free vesicles to use in control experiments. 3. Typical yields of protein incorporated into vesicles are 42 molecules of rhodopsin per vesicle, and 12 molecules of R9AP, determined by 500 nm absorbance for rhodopsin, or by densitometry of Coomassie-stained gels and comparison to a standard for R9AP. 4. The amount of accessible (i.e., cytoplasmic domain facing outward) rhodopsin and R9AP can be determined by titration with either transducin, Gαβγt-GDP, for rhodopsin, or with RGS9-1–Gβ5, for R9AP, in centrifugationbased binding assays. 5. The unilammelar character of the vesicles and their size distribution can be readily assessed by cryoelectron microscopy.

5.4 Single-Turnover GAP Assay for Purified Recombinant or Retinal Proteins To detect effects of different factors on GAP activity, two kinetic approaches have been commonly used (described in detail in Reference 30). The multiple-turnover method detects the steady-state GTP hydrolysis rate during cycles of GTPase activation and inactivation, whereas the single-turnover approach described here has been employed to avoid any possible interference caused by the G-protein recycling. A limitation of the latter approach is that it must be carried out at substrate (i.e., Gαt-GTP) concentrations well below saturating, and thus cannot be used by itself to determine values of the Michaelis-Menten constant, Km. Rather, it yields a single-exponential rate constant, kinact, which when divided by the RGS9-1 concentration yields kcat/Km, the catalytic efficiency. This value when extrapolated to in vivo conditions allows a comparison with the time constant for photoresponse recovery (figure 5.2). Protocol 6: Preparation of Urea-Washed ROS Membranes

7315.indb 106

1. Rod outer segments are prepared in the dark from fresh or frozen bovine retinas (obtained from a local abattoir, or from Schenk Packing, Seattle, Washington, or Wanda Lawson Packing, Lincoln, Nebraska) using a standard sucrose gradient procedure (29,31). 2. Each wash is carried out using the following procedure in the dark: The membrane pellet is resuspended with a 22½ gauge needle and 16 mL of wash buffer (listed in the following text). Four aliquots of 4 mL each are removed and diluted to 55 mL in a Potter-Elvehjem homogenizer (with Teflon

11/19/07 8:37:16 PM


107

GTP Hyrolyzed (percent)

100 kinact = 0.10 s–1 GST-RGS9-Gβ5 + PDE6γ

80 60

kinact = 0.045 s–1 + PDE6γ

40 20 0

0

5

10

15

20 25 Time (s)

30

35

40

Figure 5.2 GAP assay. Enhancement of GTP hydrolysis in rod outer segment membranes by added recombinant RGS9-1-G β5. The release of [32P]Pi from [γ−32P]GTP was monitored by scintillation counting of the supernatant from a charcoal suspension in samples quenched by acid at the indicated times after addition of GTP plus the indicated proteins at time zero, as described in protocol 7.

pestle) and homogenized with 10 slow strokes, care being taken not to add air bubbles to the suspension. The homogenized membranes are poured into Ti-45 centrifuge tubes (55 mL per tube), and centrifuged for 30 min at 42,000 rpm, 4°C, in a Type 45-Ti rotor (Beckman). All washes are carried out at a final concentration of 15 µM rhodopsin or below. All buffers are supplemented with 1 mM DTT and solid PMSF just before use. 3. Wash once with 1 × GAPN-Tris buffer (10 mM Tris-HCl, pH 7.4, 100 mM NaCl, 2 mM MgCl2), twice with low-salt buffer (5 mM Tris-HCl, pH 7.4, 0.5 mM MgCl2), twice with high-salt buffer (5 mM Tris-HCl, pH 7.4, 0.5 mM MgCl2, 1 M NaCl), twice with urea wash buffer (5 mM Tris-HCl, pH 7.4, 0.5 mM MgCl2, 4 M urea—urea deionized with mixed-bed ion exchange resin). The final pellet is washed once in the final assay buffer, e.g., GAPN buffer (protocol 3) for the following GAP assay (protocol 7), and after resuspension in this buffer at the desired concentration is divided into 100–200 µL aliquots wrapped in foil and stored at −80°C. Protocol 7: Single-Turnover GAP Assay

7315.indb 107

1. Urea-stripped ROS membranes are mixed with holo-transducin, Gαβγt purified from bovine retinas (see Reference 32 for purification procedure) at concentrations of 15 µM rhodopsin and 1 µM Gt in GAPN buffer (protocol 3). 2. Expose the sample to room light immediately before the assay and continuously vortex the reaction mechanically during the assay. 3. Initiate the assay by adding 50 nM [γ-32P]GTP (GTP 1 h followed by elution, and this step is repeated multiple times for Gβ5 peptide elution) or with 0.1 M glycine at pH 3.0. Eluted antigen is often of too low a concentration to be assayed by UV absorbance or by dye-binding assays, so it is usually monitored by running immunoblots on all washes and elution fractions. Usually, thrice the column volume is sufficient for efficient elution. Two to three separate washes are used for the batch procedure. 6. For analysis of coimmunoprecipitating proteins by mass spectrometry, the samples are generally loaded onto an SDS-PAGE gel and detected by Coomassie blue staining. For detection or quantification of small amounts of proteins, qualitative or quantitative (protocol 9) immunoblotting is usually used.

5.7 Localization of Proteins in Rod Outer Segments by Subcellular Fractionation and Immunofluorescence Proteins must be localized to rod or cone outer segments to play a role in phototransduction. Two useful techniques for determination and confirmation of protein localization are gradient purification of rod outer segments with each fraction quantitatively assayed for the protein of interest, and immunolocalization using fluorescent secondary antibodies and confocal microscopy.

5.7.1 Gradient Copurification: A General Way to Determine Whether a Protein Is Localized to ROS or a Contaminant from Other Parts of Retina The presence of a protein in a purified sample of rod outer segments is not sufficient to allow the conclusion to be drawn that it is a resident outer segment protein. The reason is that all purification procedures yield material that is only partially pure and always contains contamination at some level. However, it is highly unusual for a contaminating protein or organelle to display precisely the same profile across fractions of a density gradient following centrifugation as rod outer segments, especially across two different gradients, for example, one iso-osmotic gradient such

7315.indb 114

11/19/07 8:37:20 PM

Biochemical Characterization of Phototransduction RGS9-1–GAP Complex Fr. 17

18

19

20

21

22

23

24

25

26

27

28

115

29 PKCα Ab

[Rhodopsin] [PKCα]

Normalized Concentration

1 0.8 0.6 0.4 0.2 0

1

4

7

10

13

16

19

22 25

28

31

34

Fraction

Figure 5.5 Fractionation of rod outer segments, and assaying for protein comigration. ROSs were purified from bovine retinas by a discontinuous sucrose gradient, and fractions were analyzed for rhodopsin content absorbance at 500 nm and for PKCα by immunoblots.

as OptiPrep, and one sucrose gradient, which induces substantial volume loss in most organelles because of high osmolarity, but much less so for rod outer segments (figure 5.5). Protocol 12: Sucrose Gradient Fractionation

7315.indb 115

1. ROS membranes are prepared from wild-type mouse retinas in the dark by sucrose density gradient (16,29,31) or Optiprep® (iso-osmotic) density gradient (see protocol 8). 2. The best method for fractionating the gradient is by using an automatic gradient puller such as the Auto-Densi-Flow from Labconco. This instrument uses conductance to detect the liquid surface and automatically inserts the entry hole of a collection tube just below the surface. The flow into the tube (which is usually connected to a slow peristaltic pump or a slow gravityflow system) is horizontal, so there is virtually no mixing of vertical fractions. Fractions of 200 μL are collected for subsequent assays, and can be stored at −80°C until use. 3. The concentration of the major ROS marker protein rhodopsin in each fraction is determined by measuring the absorbance at 500 nm before and after light bleaching in 1.5% LDAO (N,N-dimethyldodecylamine N-oxide) detergent and 10 mM hydroxylamine. If very low amounts are used (e.g., from

11/19/07 8:37:22 PM

116


a single mouse), it may be necessary to quantify rhodopsin by quantitative immunoblotting, as described in protocol 9. 4. Proteins in these fractions are resolved by SDS-PAGE and analyzed by immunoblotting (protocol 9) to compare the purification profile of RGS9-1, Gβ5L, and R9AP or other components with that of rhodopsin. The profile of a resident ROS protein should closely follow that of rhodopsin, especially with regard to the peak position, whereas the profiles of contaminants usually only partially overlap rhodopsin’s but do not have coincident profiles.

5.7.2 Protein Localization by Immunofluorescence (12,14,19) Immunolocalization experiments are essential for visualization of protein subcellular localization, but very easily produce false results owing to antibody crossreactivity. It is useful to compare the results using different antibody preparations or using knockout mice as negative controls if possible. Useful RGS9-1 antibodies include rabbit and goat antisera raised against a C-terminal fragment of RGS9-1 (aa 226–484) (2) and a mouse monoclonal antibody that recognizes an epitope including a small part of the RGS domain and adjacent portions of the C-terminal domain (1). Protocol 13: Immunofluorescence Staining of Retinal Sections

7315.indb 116

1. Mice are humanely euthanized (e.g., by CO2 inhalation), and their eyes rapidly removed and placed in 4% paraformaldehyde in phosphate buffered saline (PBS, pH 7.2, GIBCO) for a 1 h fixation at 4°C. Some antigens require longer fixation times. 2. The eyes are cryoprotected by soaking in 30% sucrose in PBS at 4°C until tissue sinks, typically, for 6 h overnight. 3. The eyes are embedded in OCT (Tissue-Tek Compound) and rapidly frozen on dry ice or in liquid nitrogen, and stored at −80°C until use. 4. For cryosectioning, the embedded eyes are warmed to −20°C, cut into sections of 12–40 µm thickness using a cryomicrotome and placed on warm Superfrost Plus slides (Fisher). They are stored at −80°C until use. 5. Prior to staining, sections are thawed at −20°C for 1 h and 4°C for 1 h, then air dried for 30 min at room temperature. 6. Sections are dehydrated in methanol/acetone (1:1 v/v) at room temperature for 10 min. 7. Slides are washed in 0.1% Triton X-100 in PBST (PBS with 0.1% Tween 20). PBS alone may also be used; (results should be compared for each antigen) at room temperature for 2 × 10 min. 8. Sections are blocked with 10% sheep serum (Sigma) in PBST for 1 h at room temperature. 9. Slides are incubated with anti-RGS9 antibodies, anti-R9AP antiserum, or anti-Gβ5L antibody at various dilutions (1:100–1:500) in PBST (or PBS—see note in preceding paragraph) containing 10% sheep serum for 2–3 h or overnight at a chamber humidified with PBS. Wash slides with PBST for 3 × 5 min.

11/19/07 8:37:22 PM


117

10. Dye-conjugated secondary antibody is added to the slides for 1 h at the dilution recommended by the manufacturer, typically, 1:25 to 1:100 (again, it is useful when optimizing the protocol for specific antigens, antibodies, and tissues; it is best to compare different dilutions) in PBST. Then, slides are washed with PBST for 3 × 5 min. 11. The slides are mounted with a drop of Vectashield (Vector Laboratories) mounting medium and coverslipped for microscopy. Usually, the edges of the cover slip are sealed with colorless nail polish. 12. It is often useful to counterstain with a nuclear marker, such as propidium iodide, or with cell-specific markers such as peanut lectin for cone sheaths, or rhodopsin antibodies for rod outer segments.

5.8 Phosphorylation of RGS9-1 Phosphorylation is a common mechanism for regulating protein function, including RGS proteins. RGS9-1 has been reported to be phosphorylated in vitro by either protein kinase C (PKC) (16,17,36) or protein kinase A (PKA) (37) and is phosphorylated at PKC site Ser475 in vivo by a kinase whose activity is inhibited by light. In vitro phosphorylation is useful for preliminary detection of potential phosphorylation reactions and identification of sites, whereas in vivo studies using phosphorylationspecific antibodies provide insight into physiological relevance of those sites identified in vitro. Candidate kinases can be readily tested in the in vitro assays by addition of inhibitors or activators of known kinases. For these experiments, it is essential to have control substrates for the kinases in question to ensure that the effective concentrations of these substances are sufficient to activate or inhibit the endogenous kinase. For example, amphipathic inhibitors may partition into membranes, reducing their effective concentrations in solution, and some activators, such as cyclic nucleotides or diacylglycerol, may be metabolized by the cell homogenates. Protocol 14: Analysis in Vitro of Phosphorylation of RGS9-1 by Endogenous Kinases in Rod Outer Segments

7315.indb 117

1. Bovine or mouse ROSs are prepared as described earlier, typically as stocks with rhodopsin concentrations of 15–150 µm. The following procedures are all carried out in complete darkness using infrared image converters, or in dim red light. 2. Purified ROSs are homogenized as described in protocol 10 in GAPN-H buffer containing phosphatase inhibitors (and, if desired, any inhibitors or activators of specific kinases) at a dilution of 1:5, centrifuged at 8400 × g for 15 min, and then the pellet resuspended in the GAPN-H buffer to a final rhodopsin concentration of 6–60 µm. 3. NH2OH is added to the ROS at a final concentration of 10 mM to minimize rhodopsin phosphorylation, and ATP to 2–5 mM (with [γ-32P] at a specific activity of 40–100 Ci/mol if detection will be by radioactivity). The mixture is incubated at 30°C for different times up to 20 min, and the reactions stopped by washing away the free ATP buffer (by centrifuging

11/19/07 8:37:22 PM

118


+

Bovine ROS

–

ATP

Ser475-P Antibody

RGS9 Antibody

Figure 5.6 Detection of RGS9-1 phosphorylation by a phosphorylation-specific antibody. Purified bovine ROS was incubated with or without 2 mM ATP for 15 min as described in protocol 14. Proteins in ROS were analyzed by SDS-PAGE and immunoblotting with antiSer475-phosphate monoclonal antibody and anti-RGS9-1 antibody.

and resuspending the pellet thrice) with phosphatase-inhibitor buffer (5 mM Tris-HCl, 2 mM EDTA, 0.2 mM Na3VO4, 15 mM fenvalerate, 100 nM okadaic acid, 1 mM DTT). The pellets are immediately solubilized with SDS-PAGE sample buffer for detection by phospho-specific antibody, or first subjected to immunoprecipitation for detection by radioactivity. 4. ROSs are solubilized in NP-40 detergent, and RGS9-1 immunoprecipitated as described earlier and subject to SDS-PAGE, followed by autoradiography or phosphoimager analysis to detect phosphorylation. The immunoprecipitation step is critical, as RGS9-1 comigrates with tubulin in SDS-PAGE, and tubulin is a kinase substrate. 5. Alternatively, SDS-PAGE and immunoblotting with anti-Ser475-phosphate monoclonal antibody (at a dilution ratio of 1:500) are used to detect the specific Ser475 phosphorylation (see section 5.5 for immunoblottting protocol) (figure 5.6). Protocol 15: Analysis in Vivo of Phosphorylation of RGS9-1 and Regulation by Light

7315.indb 118

1. Four to six wild-type mice are maintained in a dark room for a period of 16 h, followed by euthanasia and removal of retinas under dim red light or in complete darkness with the help of infrared goggles; control mice are kept for the same time in light of a specified intensity. 2. Retinas are immediately homogenized in the dark in a 1.5 mL tube using GAPN-H buffer with 1% Nonidet P-40 detergent, plus 0.2 mM Na3VO4, 15 µm

11/19/07 8:37:23 PM


119

fenvalerate, 100 nM okadaic acid to inhibit phosphatase activities. First, a plastic pestle (Kontes) is used, and then the homogenates are sonicated on ice. 3. RGS9-1 is immunoprecipitated using rabbit polyclonal antibodies (e.g., R4432) as described earlier (protocol 10) and RGS9-1 phosphorylation is analyzed by immunoblotting with anti-Ser475-phosphate-specific antibodies following SDS-PAGE. Quantitative immunoblotting is carried out as described in section 5.5. It is difficult to obtain a standard for absolute quantification of phosphorylated RGS9-1, so usually only relative amounts (e.g., light versus dark) can be determined.

References

7315.indb 119

1. Cowan, C.W., Fariss, R.N., Sokal, I., Palczewski, K., and Wensel, T.G., Proc Natl Acad Sci U.S.A., 95, 5351, 1998. 2. He, W., Cowan, C.W., and Wensel, T.G., Neuron, 20, 95, 1998. 3. Cowan, C.W., He, W., and Wensel, T.G., Prog Nucl Acid Res Mol Biol, 65, 341, 2001. 4. Pugh, E.N., Jr., Neuron, 51, 391, 2006. 5. Snow, B.E., Krumins, A.M., Brothers, G.M., Lee, S.F., Wall, M.A., Chung, S., Mangion, J., Arya, S., Gilman, A.G., and Siderovski, D.P., Proc Natl Acad Sci U.S.A., 95, 13307, 1998. 6. Makino, E.R., Handy, J.W., Li,T., and Arshavsky, V.Y., Proc Natl Acad Sci U.S.A., 96, 1947, 1999. 7. He, W., Lu, L., Zhang, X., El-Hodiri, H.M., Chen, C.K., Slep, K.C., Simon, M.I., Jamrich, M., and Wensel, T.G., J Biol Chem, 275, 37093, 2000. 8. Kovoor, A., Chen, C.K., He, W., Wensel, T.G., Simon, M.I., and Lester, H.A., J Biol Chem, 275, 3397, 2000. 9. Simonds, W.F. and Zhang, J.H., Pharm Acta Helv, 74, 333, 2000. 10. Hu, G. and Wensel, T.G., Proc Natl Acad Sci U.S.A., 99, 9755, 2002. 11. Hu, G., Zhang, Z., and Wensel, T.G., J Biol Chem, 278, 14550, 2003. 12. Hu, G. and Wensel, T.G., Methods Enzymol, 390, 178, 2004. 13. Krispel, C.M., Chen, D., Melling, N., Chen, Y.J., Martemyanov, K.A., Quillinan, N., Arshavsky, V.Y., Wensel, T.G., Chen, C.K., and Burns, M.E., Neuron, 51, 409, 2006. 14. Zhang, X., Wensel, T.G., and Kraft, T.W., J Neurosci, 23, 1287, 2003. 15. Watson, A.J., Aragay, A.M., Slepak, V.Z., and Simon, M.I., J. Biol. Chem., 271, 28154, 1996. 16. Sokal, I., Hu, G., Liang, Y., Mao, M., Wensel, T.G., and Palczewski, K., J Biol Chem, 278, 8316, 2003. 17. Hu, G., Jang, G.F., Cowan, C.W., Wensel, T.G., and Palczewski, K., J Biol Chem, 276, 22287, 2001. 18. Chen, C.K., Burns, M.E., He, W., Wensel, T.G., Baylor, D.A., and Simon, M.I., Nature, 403, 557, 2000. 19. Lyubarsky, A.L., Naarendorp, F., Zhang, X., Wensel, T., Simon, M.I., and Pugh, E.N., Jr., Mol Vis, 7, 71, 2001. 20. Keresztes, G., Martemyanov, K.A., Krispel, C.M., Mutai, H., Yoo, P.J., Maison, S.F., Burns, M.E., Arshavsky, V.Y., and Heller, S., J Biol Chem, 279, 1581, 2004. 21. Krispel, C.M., Chen, C.K., Simon, M.I., and Burns, M.E., J Neurosci, 23, 6965, 2003. 22. Nishiguchi, K.M., Sandberg, M.A., Kooijman, A.C., Martemyanov, K.A., Pott, J.W., Hagstrom, S.A., Arshavsky, V.Y., Berson, E.L., and Dryja, T.P., Nature, 427, 75, 2004. 23. He, W. and Wensel, T.G., Methods Enzymol, 344, 724, 2002. 24. Sowa, M.E., He, W., Wensel, T.G., and Lichtarge, O., Proc Natl Acad Sci U.S.A., 97, 1483, 2000.

11/19/07 8:37:23 PM

120


25. Skiba, N.P., Martemyanov, K.A., Elfenbein, A., Hopp, J.A., Bohm, A., Simonds, W.F., and Arshavsky, V.Y., J Biol Chem, 276, 37365, 2001. 26. Martemyanov, K.A., Lishko, P.V., Calero, N., Keresztes, G., Sokolov, M., Strissel, K.J., Leskov, I.B., Hopp, J.A., Kolesnikov, A.V., Chen, C.K., Lem, J., Heller, S., Burns, M.E., and Arshavsky, V.Y., J Neurosci, 23, 10175, 2003. 27. Baker, S.A., Martemyanov, K.A., Shavkunov, A.S., and Arshavsky, V.Y., Biochemistry, 45, 10690, 2006. 28. Litman, B.J., Methods Enzymol, 81, 150, 1982. 29. Papermaster, D.S. and Dreyer, W.J., Biochemistry, 13, 2438, 1974. 30. Cowan, C.W., Wensel, T.G., and Arshavsky, V.Y., Methods Enzymol, 315, 524, 2000. 31. Papermaster, D.S., Methods Enzymol, 81, 48, 1982. 32. Wensel, T.G., He, F., and Malinski, J.A., Methods Mol Biol, 307, 289, 2005. 33. Gill, S.C. and von Hippel, P.H., Anal Biochem, 182, 319, 1989. 34. Liang, Y., Fotiadis, D., Filipek, S., Saperstein, D.A., Palczewski, K., and Engel, A., J Biol Chem, 278, 21655, 2003. 35. Bradford, M.M., Anal Biochem, 72, 248, 1976. 36. Nair, K.S., Balasubramanian, N., and Slepak, V.Z., Curr Biol, 12, 421, 2002. 37. Balasubramanian, N., Levay, K., Keren-Raifman, T., Faurobert, E., and Slepak, V.Z., Biochemistry, 40, 12619, 2001.

7315.indb 120

11/19/07 8:37:24 PM

6

Guanylate CyclaseBased Signaling in Photoreceptors and Retina Karl-Wilhelm Koch and Andreas Helten

Contents 6.1

Introduction.................................................................................................. 121 6.1.1 Guanylate Cyclases in the Retina..................................................... 122 6.2 Membrane-Bound Photoreceptor Guanylate Cyclases................................ 122 6.2.1 Guanylate Cyclase Assays................................................................ 124 6.2.2 Native Guanylate Cyclase from Purified ROS.................................. 125 6.2.3 Recombinant ROS-GCs.................................................................... 129 6.3 Guanylate Cyclase-Activating Proteins....................................................... 130 6.3.1 Complex of GCAP1 and GCAP2 with ROS-GC1............................ 131 6.3.2 Analysis of GCAP Properties........................................................... 133 6.3.3 Chemical Modification of GCAPs.................................................... 135 6.3.4 Reconstitution of Purified GCAPs with Membrane-Bound ROS-GCs.......................................................................................... 136 Acknowledgments................................................................................................... 138 References............................................................................................................... 139

6.1 INTRODUCTION The cyclic nucleotides cyclic AMP (cAMP) and cyclic GMP (cGMP) function as a second messenger in different tissues such as brain, heart, kidney, lung, eye, nose, and smooth and skeletal muscle. Cyclic nucleotides are synthesized from ATP or GTP by different isoforms of an adenylate cyclase (AC) or guanylate cyclase (GC), respectively. Nine membrane-bound and one soluble AC have been described in mammals so far, and some of these isoforms show an ubiquitous tissue distribution (1). Mammalian GCs exist also in particulate and soluble forms; they are classified in 121

7315.indb 121

11/15/07 1:40:00 PM

122


seven membrane-bound GCs (GC-A to GC-G) and six soluble GC-subunits (α1-3 and β1-3) (2–6). Membrane-bound GCs form homodimers, whereas soluble GCs operate as heterodimers consisting of an α- and β-subunit. Membrane GCs are integrated into the membrane by one transmembrane region. By this arrangement, they are activated by extracellular signaling molecules (e.g., hormones) and transmit the primary signal to subsequent intracellular steps. Whereas three GC forms operate in this manner and function as either hormone receptors, targets for bacterial enterotoxins, or targets for intestine-derived small peptides (guanylin), other forms are named orphan receptors, because no extracellular ligands that bind and regulate these GCs have been identified so far. However, the term orphan receptor is misleading for at least two of these membrane-bound GCs, which are expressed in photoreceptor cells of the mammalian retina (7–9). They are regulated by small Ca2+-binding proteins on their cytoplasmic domains (see following text). Targets of cyclic nucleotides in cell signaling include cAMP- and cGMP-dependent protein kinases (PKA and PKG) (10,11), cyclic nucleotide-gated (CNG) cation channels (12,13), phosphodiesterases (PDE) (14,15), and guanine-nucleotide-exchange factors (epac, exchange protein directly activated by cAMP) (16). Synthesis of cyclic nucleotides by cyclases is counterbalanced by the activity of a class of PDEs.

6.1.1 Guanylate Cyclases in the Retina The present chapter focuses on GCs in the vertebrate retina, where membrane-bound (7–9) and soluble GCs have been found and localized (17–22). Photoreceptor cells express a subset of particulate GCs named ROS-GC1 and ROS-GC2 or, alternatively, RetGC1, RetGC2, or GC-E and GC-F (7–9). The most characteristic and distinguishable feature of these GCs is that they do not respond to hormone peptides, but instead are regulated by small Ca2+-binding proteins called GCAPs (guanylate cyclase-activating proteins) on their intracellular site (7–9;23–25). Although the presence of natriuretic peptides and their receptor GCs in the retina has been described for several species (20,26,27), much less is known about their specific function in this tissue. Soluble GCs are localized in several layers of the retina (17–22) and are the main target of the gaseous messenger nitric oxide (NO).

6.2 Membrane-Bound Photoreceptor Guanylate Cyclases Vertebrate photoreceptor cells respond to light with a hyperpolarization of their plasma membrane. The capture of photons by the visual pigments (rhodopsin and cone opsins) triggers a G-protein-coupled enzymatic cascade that controls the cytoplasmic level of the internal messenger cGMP in the outer segments of photoreceptor cells (28,29). CNG channels in the plasma membrane are opened by high cytoplasmic concentrations of cGMP. This so-called dark state of the cell is characterized by a constant flow of Na+ and Ca 2+ ions into the outer segment of the cell. Flow of Ca2+ into the cell is balanced by continuous extrusion via a Na+:Ca2+, K+exchanger that is also located in the plasma membrane. Light-induced hydrolysis

7315.indb 122

11/15/07 1:40:01 PM

123

Guanylate Cyclase-Based Signaling in Photoreceptors and Retina

of cGMP leads to the closure of the CNG-channels, which stops the ion flow into the cell. However, the exchanger continues to operate and expels Ca 2+ out of the cell, leading to a net decrease of cytoplasmic Ca 2+ concentration ([Ca 2+]). Recovery of the cell from illumination and hyperpolarization requires the shutoff of all exciting steps in the transduction cascade; in addition, it requires the refilling of the exhausted cGMP pool. GCs catalyze the synthesis of cGMP from GTP in outer segments under control of a negative Ca 2+ feedback loop: (7–9,28,29) decreasing Ca2+ increases GC activity and vice versa. The action of Ca 2+ on GC activity is not direct, but is mediated by small Ca 2+-binding proteins dubbed GCAPs (23–25) (figure 6.1). GCAPs belong to a group of neuronal calcium sensor (NCS) proteins (25). They are related to another Ca2+ sensor named recoverin, which is also part of a negative feedback loop (figure 6.1). Regulation of ROS-GCs by GCAPs has been studied by different approaches using (a) native rod outer segment (ROS) preparations, (b) membranes prepared from ROS in combination with recombinant GCAPs, or (c) recombinant GCs reconstituted with recombinant GCAPs. Rec 2+

Ca RhK

Rec

RhK Rh

Rh

–Ca2+ +Ca2+

Ca2+

GCAP-1

GCAP-2

GCAP-1

GC

GTP cGMP +PPi

GCAP-2

GC

GTP

cGMP +PPi

Figure 6.1 Ca2+ feedback reactions in photoreceptor cells involving NCS proteins. Disk membranes in vertebrate photoreceptor cells harbor a high density of the photopigment rhodopsin (Rh) and other membrane-bound proteins, including photoreceptor guanylate cyclases (ROS-GC1 and ROS-GC2). Membrane-bound GCs form dimers and are regulated by small Ca2+-binding proteins named GCAPs. The Ca2+ concentration in the dark state of the cell is high and keeps the cyclase activity at a basal rate (left part of figure). Illumination leads to a decrease of cGMP and cytoplasmic [Ca2+], which in turn triggers several Ca2+-dependent feedback reactions. For example, a rearrangement of the GCAP–GC complex leads to an increase of cyclase activity and a higher rate of cGMP synthesis. GCAPs are related to another NCS protein named recoverin (Rec). Recoverin inhibits rhodopsin kinase (RhK) at high [Ca2+] and thereby prevents phosphorylation of rhodopsin. Decreasing [Ca2+] triggers a Ca2+-myristoyl switch in recoverin that facilitates the detachment of recoverin from the membrane and relieves inhibition of rhodopsin kinase.

7315.indb 123

11/15/07 1:40:02 PM

124


6.2.1 Guanylate Cyclase Assays Guanylate cyclase activity can be assayed by different methods; the principle of each method is described briefly in the following text:

1. Conversion of [α-32P]GTP to [α-32P]cGMP (30,31). The radiolabeled nucleotides are separated by thin-layer chromatography, spots containing nucleotides are identified and cut out, and radioactivity is counted in a scintillation counter. The loss of cGMP by hydrolysis due to PDE activity can be corrected by inclusion of [3H]cGMP as an internal standard. 2. A radiolabeled thio-analog of GTP, (Sp)-GTPαS, is used as substrate (32). ROS-GCs produce (Rp)-cGMPS from (Sp)-GTPαS by a cyclization reaction that involves the inversion of the configuration at the α-phosphorus atom (33). Hydrolysis of (Rp)-cGMPS by the photoreceptor PDE is very low, which makes this assay suitable for operation in the presence of activated PDE. Separation of nucleotides is performed on aluminum oxide (32). 3. The cyclization reaction performed by GCs yields, as a second reaction product, pyrophosphate (PPi). A spectrophotometric assay measures the production of PPi by employing several enzyme-coupled reactions that finally lead to the oxidation of β–NADH (34). 4. Classical radioimmunoassays determine the amount of cGMP produced from a competition experiment with a radiolabeled cGMP standard (35). 5. Nucleotides can be separated on a reverse-phase high-performance liquid chromatography (HPLC) column (31,33,36). Eluted peaks are detected at 254 nm, and amounts of formed cGMP are calculated from the peak area by referring to a calibration curve of precisely known cGMP standards. The assay has the advantage that it does not use radiolabeled compounds. Modern HPLC systems allow operation with high sensitivity and automatic sample injection. The detection limit can be as low as 5 pmol nucleotide. Large sample numbers can be processed in overnight runs. Working with ROS or ROS membranes, however, requires the use of specific PDE inhibitors (e.g., zaprinast) and the performance of the assay in complete darkness or under infrared illumination. Protocol 1: HPLC Assay of Guanylate Cyclase Activity Samples containing ROS-GC are incubated with GC-assay buffer. The free [Ca2+] in the sample mix is adjusted and varied with different Ca2+-EGTA buffers. Reaction products are analyzed by HPLC. The HPLC system consists of two pumps, a detector, and an autosampler and is controlled by commercial software provided by the manufacturer.

7315.indb 124

1. Prepare a GC-assay buffer stock solution (2.5 × Mg2+ GC buffer) containing 100 mM Hepes-KOH, pH 7.5, 140 mM KCl, 20 mM NaCl, 25 mM MgCl2, 5 mM GTP, 1 mM zaprinast, 0.25 mM ATP). In cases where you plan experiments with Mn2+ instead of Mg2+, substitute 25 mM MgCl2 for 5 mM MnCl2.

11/15/07 1:40:02 PM


125

2. Prepare Ca2+-EGTA buffer using stock solutions of K2CaEGTA and K2H2EGTA. A detailed description of the preparation of these stock solutions has been provided elsewhere (37). Mix the stock solutions at different ratios by always keeping the total EGTA concentration constant at 2 mM. Use a calcium buffer program (for example, WEBMAXC Standard: http:// www.standford.edu/~cpatton/webmanxcS.htm) to calculate the free [Ca2+] in the incubation mixture. Check the free [Ca2+] in the mixtures with fluorescent Ca2+ indicators and/or a Ca2+ electrode as described (31,38–40). 3. Mix 20 µL of the GC assay buffer with 10 µL of bidistilled water and with 10 µL of Ca2+ EGTA buffer of desired free [Ca2+]. If necessary, 10 µL of water can be substituted by other ingredients that undergo a test concerning their influence on GC activity (e.g., inhibitors, peptides, and others). GC activity is further increased by a final concentration of 100 µM ATP in the assay mixture. 4. Turn off the room light. Use only indirect and very dim red light. Keep the ROS sample in complete darkness. Start reaction by addition of 10 µL ROS, and incubate for 5 min at 30°C. Alternatively, the reaction can be started by addition of 20 µL of GC assay buffer to a premix of the other components. If this route is chosen, preincubate the premix for 5 min at room temperature before starting the reaction. Terminate the reaction by adding 50 µL of ice-cold 100 mM EDTA and heat the samples for 5 min at 95°C. All other steps can be performed under daylight conditions. Vary the incubation time according to your specific aims. 5. Centrifuge the reaction samples for 5 min at 14,000 × g in a table centrifuge. Remove carefully 50–90 µL of the supernatant. The exact volume depends on the injection syringe and injection loop of your HPLC system. Transfer the supernatant of your samples to appropriate HPLC vials. 6. Prepare 2 L of the HPLC running buffer A containing 5 mM KH2PO4; the pH will be around 5.0, and further adjustment of the pH is not necessary. Filtrate the buffer with a filtration device using a hydrophilic polypropylene filter (0.22 µm). Filtrate also methanol (1 L). 7. For separation of nucleotides use a reverse-phase column such as LiChro CART 250-4, LiChrospher 100 RP-18 (5 µm) provided by Merck. Set the flow rate to 1.2 mL per minute. Equilibrate the column in buffer A (5 mM KH2PO4). Start the run by injection of the sample. Elution of nucleotides is performed with a two-step gradient: Increase methanol from 0 to 15 % within 4 min. Then, increase methanol from 15 to 70% within 5 min. Decrease methanol from 70 to 0 % within 1 min. Reequilibrate for 4 min with buffer A before the next injection cycle can start. 8. Identify the cGMP peak in the chromatogram (figure 6.2), and determine the area of the peak with appropriate software.

6.2.2 Native Guanylate Cyclase from Purified ROS A membrane-bound guanylate cyclase was originally purified from bovine, toad, and frog ROS by fractionation studies in combination with activity measurements (36,41). The isolated protein from bovine ROS had a molecular mass of 110–112 kDa

7315.indb 125

11/15/07 1:40:02 PM

126

Signal Transduction in the Retina GTP

GDP cGMP G

0

2

4

6 Time (min)

8

10

Figure 6.2 HPLC chromatogram of a nucleotide mixture obtained after assaying ROS-GC activity. A mixture of purified ROS was incubated with GC assay buffer at a free [Ca2+] of 127 nM according to protocol 1. Separation of nucleotides on a reverse-phase column yielded a cGMP peak at a retention time of 6.1 min. The peak area is linearly related to the amount of cGMP (range 5–5000 pmol) and can be calculated from a calibration curve. The GC substrate GTP is also hydrolyzed by the GTPase activity of transducin, yielding GDP. In addition, nonspecific nucleotidases present in ROS produce guanosine (G) from GTP and GDP.

and was present in fractions with the highest specific activities (36), and up to five polymorphic variants of a membrane-bound GC were isolated from amphibian retinae (41). Subsequent partial amino acid sequencing of the isolated proteins (42,43) revealed their identity with a cDNA clone obtained from human genomic DNA (44) and a bovine retina cDNA library (45). Shortly after identification of this GC (thereafter named ROS-GC1, retGC1, or GC-E) (46–48), the cloning of a second form of a photoreceptor GC was reported for several vertebrate species (named ROSGC2, retGC2, or GC-F) and its localization in photoreceptor cells was demonstrated (23,49). However, so far this protein has not been purified from native sources, possibly because it is only present in low amounts in bovine retina, the main starting material for the isolation of retinal proteins. Although the precise ratio of both GCs remains to be determined and their contribution to the overall synthesis of cGMP in rod and cone cells is still a matter of discussion (9), one can use purified bovine ROS as a source to enrich and purify the isoform ROS-GC1 (36). Protocol 2: Solubilization and Purification of ROS-GC1 Starting materials for purification of ROS-GC1 are bovine ROS prepared from dark-adapted bovine retinae that were freshly obtained from a local slaughterhouse. Soluble cytoplasmic proteins are removed first by washing in hypoosmotic buffer. Washed membranes are then solubilized in detergent, and proteins are selectively extracted by buffers that differ in detergent and salt content. Rhodopsin and other membrane proteins are removed first by the Triton buffer. Rhodopsin is seen as the prominent protein band in the TritonX-100 extract (TE) in figure 6.3a. ROS-GC1 is solubilized in a subsequent extraction

7315.indb 126

11/15/07 1:40:03 PM


127

TE DS DE 2nd DE kD

GC Activity (nmol × mg–1 × min–1)

200 160

250 150 100 75

ROS-GC1

50 37

120

25 20

80 40 0

DS

DE

2nd DE

Figure 6.3a Solubilization of ROS-GC1 from ROS membranes. Specific guanylate cyclase activities of fractions DS, DE, and 2nd DE (see following text). Fractions were obtained by a selective detergent extraction procedure and analyzed by SDS-PAGE (inset: approx. 1 µg protein in lanes TE to 2nd DE). TE, ROS membrane proteins extracted by TritonX-100 and low salt; DS, suspension of solubilized ROS proteins after removal of proteins that are soluble in the Triton-buffer (TE); DE, n-dodecyl-β-d-maltoside extract obtained from DS after centrifugation; 2nd DE, extract obtained after a second extraction step with n-dodecyl-β-dmaltoside. A molecular weight marker is shown on the left side.

step using n-dodecyl-β-d-maltoside as detergent (DE in figure 6.3a). Repeating this extraction step leads to an already highly enriched fraction of ROS-GC1 (2nd DE in figure 6.3a). Purified ROS-GC1 is obtained after chromatography on an anion exchange column and a GTP-agarose-affinity column (36). Alternatively, ROS-GC1 can be purified on lectin columns such as Concanavalin A-Sepharose or wheat germ agglutinin (WGA)-Sepharose (43).

7315.indb 127

1. Use 4 mL of purified ROS with a rhodopsin concentration between 6 and 10 mg/mL. Scale up, if necessary. Work in dim red light. 2. Dilute ROS with low-salt buffer (10 mM Hepes-KOH, pH 7.4, 1 mM DTT, 0.1 mM PMSF, 0.1 mM EGTA) about fivefold; use, for example, four vials and centrifuge for 30 min at 50,000 × g and at 4°C. Remove the supernatant and repeat the washing step with low-salt buffer. 3. Remove the supernatant after the last washing step, and resuspend every pellet in 1 mL of Triton buffer (5% [v/v] Triton X-100, 2 mM MgCl2, 10 mM HepesKOH, pH 7.4, 0.1 mM PMSF, 0.1 mM EGTA, 1 mM DTT). Combine all suspensions, and adjust a final concentration of 4 mg rhodopsin per milliliter by adding Triton buffer. All subsequent steps can be performed in daylight. 4. Homogenize the suspension in a homogenizer (e.g., Elvehjem homogenizer). Keep the homogenized suspension on ice for less than 30 min. Centrifuge

11/15/07 1:40:04 PM

128


the suspension for 20 min at 165,000 × g at 4°C. Microultracentrifuges are most convenient at this step. 5. Remove the supernatant and resuspend the small pellet in DM-buffer (20 mM n-dodecyl-β-d-maltoside, 20 mM Hepes-KOH, pH 7.4, 1 M NaCl, 0.1 mM PMSF, 0.1 mM EGTA, 1 mM DTT). Adjust to 4 mg rhodopsin per milliliter (corresponding to the initial amount of rhodopsin in the starting material), and homogenize the suspension. Keep on ice for 1 h. Centrifuge for 15 min at 100,000 × g at 4°C. 6. Remove the supernatant and keep at −80°C until use. This supernatant contains an enriched fraction of photoreceptor GC. However, the small pellet also contains GC and, therefore, the extraction procedure is repeated. Use a total of 1.6 mL DM buffer, when your starting material was 4 mL ROS. Collect the supernatant after centrifugation. 7. The second DM extract prepared as described in step 6 is already highly enriched in ROS-GC1. It can be purified further by chromatography on a WGA-Sepharose column (column volume 1–2 mL). Adjust the second DM extract to the following buffer using either ultrafiltration or a spin column: 2 mM n-dodecyl-β-d-maltoside, 25% glycerol, 20 mM Tris-HCl, pH 7.4, 100 mM NaCl, 2 mM MnCl2, 1 mM CaCl2, and 1 mM DTT. Equilibrate the WGA-Sepharose with 20 mM Tris, pH 7.4, 2 mM MnCl2, 1 mM CaCl2 100 mM KCl, 25% glycerol, and 1 mM DTT. Incubate the ROS-GC-containing solution with the WGA-Sepharose for 1 h. Remove the unbound material, wash the WGA-Sepharose by 3–4 column volumes of equilibrating buffer.

GC-activity (nmol cGMP × min–1 × ml–1)

10

11

12

13

2.0

1.5

1.0

0.5

0.0 0

2

4

6 8 Volume (ml)

10

12

14

Figure 6.3b Purification of ROS-GC1 on WGA-Sepharose. Guanylate cyclase activities were measured in each fraction. Elution of bound ROS-GC1 by n-acetyl-d-glucosamine starts at an elution volume of 10 mL. Inset, SDS-Page analysis of fractions with peak of activity; see arrow that indicates ROS-GC1. Protein bands were stained with silver.

7315.indb 128

11/15/07 1:40:05 PM


129

Elute-bound ROS-GC1 with the following buffer: 20 mM n-dodecyl-β-dmaltoside, 20 mM Hepes-KOH, pH 7.4, 1 M NaCl, 1 mM DTT, and 0.5 M n-acetyl-d-glucosamine. The activity profile of fractions obtained during chromatography on WGA-Sepharose is shown in figure 6.3b. Purified ROS-GC1 was eluted from the column in fractions 10–13 mL as a protein band at 112 kDa (see arrow in figure 6.3b); its identity was verified by a ROS-GC1 specific antibody (43). The specific activities of purified ROSGC1 preparations (36,43) were at least 240 nmol cGMP × mg−1 × min−1. ROS-GC2 was not detectable in these fractions as tested by a ROS-GC2 specific antibody.

6.2.3 RECOMBINANT ROS-GCs Heterologous expression of ROS-GCs in cell culture is widely used to study regulatory mechanisms by GCAPs or to investigate the biochemical consequences of site-directed mutagenesis of the ROS-GC1 or ROS-GC2 gene. COS cells, HEK293 cells, or tsA-cells (a modified version of HEK293 cells) are used for most applications (39,40,45,49). Protocol 3: Heterologous Expression of ROS-GCs in tsA Cells

1. Cultivate cells in medium M10 (minimal essential medium with 10% fetal calf serum, 2 mM l-glutamin, 1% nonessential amino acids, and 1% antibiotika/antimyotika; GIBCO BRL) at 37°C, 5% (p/p) CO2, and about 95% humidity on petri dishes. 2. Seed 3 × 105 cells per 5 cm dish or 9 × 105 cells per 9 cm dish. Change medium (M10) after 22 h. Wait for additional 1 to 2 h before start of transfection. 3. Vector constructs pcDNA3.1-GC1 or pcDNA3.1-GC2 containing coding regions of ROS-GC1 or ROS-GC2, respectively, were used for gene transfer into cells by the calcium phosphate method, according to Chen and Okayama (50). Dissolve 30 µg of DNA in 372 µL of H2Obidest. Add 123 µL 1 M CaCl2 and 495 µL 2 × BBS-buffer (50 mM BES, 280 mM NaCl, 1.5 mM Na2HPO4, pH 6.95). Mix and incubate for 20 min at room temperature. Pipette this solution to a 9 cm petri dish containing cells in culture medium. Incubate for 20–22 h at 35°C under 3% (p/p) CO2 and approximately 95% humidity. 4. Remove the DNA/calcium phosphate precipitates by washing with 5 mL PBS (phosphate-buffered saline) buffer and subsequently with 3 mL PBS/0.5% (w/v) EDTA and 5 mL PBS. Refill dishes with 8 mL of prewarmed M10 medium. Incubate for 20–22 h at 37°C, 5% CO2, and 95% humidity until harvesting.

Cell membranes of tsA cells transfected with DNA of ROS-GCs can be used for further studies on guanylate cyclase regulation. For this purpose, cell membranes need to be prepared from tsA cells.

7315.indb 129

11/15/07 1:40:05 PM

130


Protocol 4: Preparation of tsA Cell Membranes Containing Heterologously Expressed ROS-GC

1. Collect tsA cells in medium from petri dishes (9 cm) and pellet by centrifugation (200 × g; 5 min, 4°C). Resuspend cells in 8 mL PBS buffer and repeat centrifugation. Resuspend the cell pellet in 100 µL lysis buffer (10 mM Hepes-KOH, pH 7.5, 1 mM DTT). At this stage cells can be shockfrozen with liquid nitrogen and stored at −80°C. 2. Sonify a thawed cell suspension for 4 × 5 s (Branson Sonifier B12, 80–100 W). Centrifuge lysed cells (400 × g, 5 min). 3. Take the supernatant and pellet the membranes by centrifugation in an ultracentrifuge (125,000 × g, 15 min, 4°C). Resuspend the membrane pellet (corresponding to cells from a 9 cm dish) in 100 µL 10 mM Hepes-KOH, pH 7.5, 250 mM KCl, 10 mM NaCl, 1 mM DTT. Incubate for 30–60 min. Homogenize by short sonification (5 s). Determine the protein concentration using standard methods (e.g., Amido Black method). Typically, 100–200 µg protein was obtained from a 9 cm petri dish resulting in a final protein concentration of tsA membranes of 1–2 mg/mL.

6.3 GUANYLATE CYCLASE-ACTIVATING PROTEINS Outer segments of vertebrate rod and cone cells harbor a set of Ca2+-binding proteins that operate as Ca2+ sensors during a light response. They detect changes in intracellular [Ca2+] depending on the illumination conditions and regulate their targets in a Ca2+-dependent fashion. Recoverin inhibits rhodopsin kinase (GRK1) at high [Ca2+] in a dark-adapted cell (51). Inhibition is relieved when the [Ca2+] drops after illumination (figure 6.1). Calmodulin binds to the CNG channel at high [Ca2+] and dissociates from the binding sites when [Ca2+] decreases (52,53). This step increases the affinity of cGMP for the channel, thereby facilitating the reopening of the channel. Guanylate cyclase-activating proteins (GCAPs) activate ROS-GCs at low [Ca2+] and show a modest inhibition below the basal GC activity level at high [Ca2+] (23–25). The Ca2+-dependent regulation of ROS-GCs by GCAPs is one of the main negative feedback loops that adjust the light sensitivity of a photoreceptor cell during illumination. A typical example of guanylate cyclase activity as a function of the [Ca2+] is shown in figure 6.4. Mammalian photoreceptor cells express up to three, zebra fish photoreceptors up to six, and pufferfish photoreceptors up to eight different GCAP isoforms (55). Most functional studies in the past, however, were restricted to mammalian GCAP1 and GCAP2. These proteins were originally isolated from mammalian rod outer segments or retinae preparations (56–58), but their easy and reliable expression in E. coli cells allowed a large range of functional studies with recombinant proteins. Previous work included, but was not limited to, the identification and characterization of target sites in ROS-GC1 (59–62), structure–function studies on GCAP mutants (63–68), and the characterization of disease-related mutations in ROS-GC1

7315.indb 130

11/15/07 1:40:06 PM


131

GC Activity (nmol × min–1 × mgRh–1)

10 8 6 4 2 0

0.1

1

10

[Ca2+] (µM)

Figure 6.4 Ca2+-dependent activation of ROS-GC in whole ROS. Typical profile of ROS-GC activities in a range of free [Ca2+] from 50 nM to 10 µM. Half-maximal activation (usually expressed as IC50) was at 237 nM free [Ca2+].

and GCAP1 (23,69). Recordings from transgenic mice with altered expression levels of GCAP1 and GCAP2 confirmed the principal operational mechanism of Ca2+dependent regulation of ROS-GCs (70–73). Details of the foregoing aspects have been broadly covered in several recent reviews (8,9,23–25) and will not be discussed here. Instead, we will give a short summary of the main regulatory features of ROSGC1 controlled by the action of GCAP1 and GCAP2. ROS-GC2 is not covered, because most previous work has focused on ROS-GC1.

6.3.1 Complex of GCAP1 and GCAP2 with ROS-GC1 Results from independent experimental approaches indicated that ROS-GC1 and GCAPs form a complex at low and high [Ca2+] (62,63,74,75). In the Ca2+-loaded form of GCAP, i.e., the resting state, the interaction between GCAP and GC leads to a conformation of GC that is only able to synthesize cGMP at a low rate. Increase in cGMP synthesis rate is triggered by a conformational change in GCAP1 and/or GCAP2, which subsequently leads to a different interaction modus operandi. The increase in GC catalytic activity is probably enabled by a better orientation of the catalytic domains in the GC dimer. Details of the activation mechanism are still missing and are a matter of current debate. For example, a “priming effect” of ATP has been suggested to be a necessary condition for ROS-GCs to achieve high levels of physiologically relevant enzymatic activities (76–78). Another line of recent research proposes that Mg2+, not Ca2+, is the physiologically important cation to be associated with GCAPs (39). Heterologously expressed and purified GCAPs are useful tools to test these working models.

7315.indb 131

11/15/07 1:40:07 PM

132


Protocol 5: Heterologous Expression of GCAPs

1. Use Escherichia coli BL21-Codon Plus (DE3) cells for overexpression of GCAPs. Transform competent E. coli cells with plasmids containing GCAP-DNA. Use pET-11a/GCAP1 for expression of wild-type GCAP1 and pET21a/GCAP2 for expression of wild-type GCAP2. Use the plasmid pBB131 containing the gene for a yeast N-myristoyl-transferase (kindly provided by Dr. J. Gordon, Washington University School of Medicine, St. Louis, Missouri), when myristoylated variants of GCAPs are produced. Use a D6S mutant of GCAP1 in order to maximize degree of myristoylation. This mutation is not necessary for the heterologous expression of myristoylated GCAP2, because the amino acid sequence of GCAP2 harbors a complete consensus site for myristoylation by yeast N-myristoyl-transferase. 2. Thaw 100 µL of competent cells on ice, add plasmid DNA (1–10 ng) and incubate for 30 min on ice. Apply a heat pulse of 42°C for 25 s, and incubate on ice for 2 min. Add 400 µL of LB medium and grow cells for one hour at 37°C. Spread 50–500 µL on agar plates containing appropriate antibiotics (100 µg/ mL ampicillin and 30 µg/mL kanamycin). Incubate overnight at 37°C. 3. Inoculate 5 mL of dYT medium supplemented with 100 µg/mL ampicillin and 25 µL/mL kanamycin with single colonies of transformed cells and grow at 37°C overnight. Add 2 mL of an overnight culture to 500 mL of dYT medium containing 100 µg/mL ampicillin and 25 µg/mL kanamycin. Grow by vigorous agitation at 37°C. Add myristic acid (100 µg/mL final concentration, dissolved in ethanol) at an OD of 0.4 (in case you plan to express a myristoylated protein). Induce expression at an OD of 0.6 by adding isopropyl-β-d-thiogalactoside (IPTG, 1 mM final concentration). Harvest cells after 4 h by centrifugation (5000 × g, 10 min, 4°C). Resuspend the pellet of 1 L bacterial culture in 40 mL of 50 mM Tris-HCl, pH 8.0. 4. Add 100 µg/mL lysozyme and 5 u/mL DNAse to the bacteria suspension. Incubate at 30°C for 30 min (water bath), and add 1 mM DTT and 0.1 mM PMSF. Centrifuge at 360,000 × g for 30 min at 4°C. GCAPs are partially present in the soluble fraction, but they are also present in high amounts in inclusion bodies. For further purification from the soluble fraction, continue at protocol 6. 5. Solubilize the pellet in 6 M guanidinium-hydrochloride, 1 mM DTT (40–60 mL per 1 L of culture). Solve by stirring at room temperature. Dissolve against 5 L of dialysis buffer (150 mM NaCl, 20 mM Tris-HCl, pH 8.0, 1 mM DTT) for 5 h. Repeat dialysis. 6. Precipitate insoluble material by centrifugation (360,000 × g, 30 min, 4°C). Take the supernatant and add 67% (NH4)2SO4 to precipitate proteins. Pellet by centrifugation at 60,000 × g for 30 min at 4°C. The protein pellet can be stored at −20°C for later use. Protocol 6: Purification of GCAPs Purification of GCAP is achieved by two chromatographic steps, first, by size exclusion chromatography (SEC) and, second, by ion exchange chromatography

7315.indb 132

11/15/07 1:40:07 PM


133

(IEC) (58,75,79). If GCAPs were obtained from the soluble bacterial fraction, concentrate the protein solution to the appropriate volume before applying onto the SEC column.

1. Dissolve the (NH4)2SO4 pellet of expressed GCAP in 2–3 mL of bidistilled water. Remove undissolved material by centrifugation (100,000 × g, 15 min, 4°C). Equilibrate a HiLoad 16/60 Superdex prep grade column with gel filtration buffer (150 mM NaCl, 20 mM Tris-HCl, pH 7.5, 1 mM DTT). Add 2 mM EGTA for the purification of GCAP1 or 2 mM CaCl2 for the purification of GCAP2. Adjust to a flow rate of 1 mL/min. Inject the GCAP sample. Collect the eluted fraction, and test for presence of GCAP by SDS-PAGE. 2. Combine the GCAP-containing fractions and adjust to 50 mM NaCl by dilution with 20 mM Tris-HCl, pH 7.5, 1 mM DTT. Equilibrate an IEC column (MonoQ or UnoQ) with IEC buffer A (50 mM Tris-HCl, pH 7.5, 50 mM NaCl, 1 mM DTT, EGTA, or CaCl2 as indicated earlier). Set the flow rate to 0.5 mL/min, and apply the protein on the column. Wash the nonbound proteins from the column. Set the flow rate to 3 mL/min, and elute bound proteins by a gradient of 13 column volumes of 0–55% buffer B (20 mM Tris, pH 7.5, 1 M NaCl, 1 mM DTT, EGTA, or CaCl2 as indicated). Collect the fractions and analyze by SDS-PAGE.

6.3.2 Analysis of GCAP Properties GCAPs constitute a subgroup of neuronal calcium sensor (NCS) proteins. These proteins bind Ca2+ with moderate to high affinity and exhibit Ca2+-induced conformational changes. Direct Ca2+ binding can be investigated by radioisotope methods using 45Ca2+ or by employing fluorescent indicators to monitor free [Ca2+] in titration experiments (39,63,64,68). A simple and quick test for Ca2+-induced conformational changes in GCAPs is a gel-shift assay (58,68,79). GCAP samples were complemented with either 1 mM CaCl2 or 1 mM EGTA and were analyzed by SDS-PAGE. Protein bands of GCAP1 and GCAP2 exhibit a large change in their electrophoretic mobility. The apparent molecular weights of the Ca2+-bound and Ca2+-free forms are 20–21 kDa and 25–26 kDa, respectively. This simple method is a qualitative test to determine whether the Ca2+-induced conformational change in a GCAP mutant is still taking place or is significantly disturbed. Another method that is widely used to study Ca2+-binding proteins is to record tryptophan fluorescence as a function of [Ca2+] (39,63,68). Many Ca2+-binding proteins exhibit a Ca2+-dependent change in the maximum of the tryptophan fluorescence emission. These changes reflect changes in the conformation of the protein that are monitored by tryptophan residues. The protein sample is excited at 280–290 nm, and the emission spectrum is recorded between 300 and 450 nm. Decreasing the [Ca2+] from 10 −3 to 10 −9 M changes tryptophan fluorescence emission at 335 nm. The relative fluorescence emission at maximum is then plotted as a function of the [Ca2+]. An example of a Ca2+-dependent change in tryptophan fluorescence of GCAP1 is shown in figure 6.5. The Ca2+-dependent tryptophan fluorescence emission of GCAP1 is biphasic (39,63,68).

7315.indb 133

11/15/07 1:40:07 PM

134


1

Fluorescence (arbitrary units)

60000

2 3 4

40000

20000

0 300

320

340 360 380 Wavelength (nm)

400

420

Figure 6.5 Tryptophan fluorescence of nonmyristoylated GCAP1. The tryptophan fluorescence emission of 1.1 µM GCAP1 was recorded in the presence of different [Ca2+]. Trace 1 was obtained in 2 mM K 2H2EGTA (no K 2CaEGTA added). Free Ca2+ concentrations in the other traces was 5 nM (trace 2), 11 nM (trace 4), and 10 µM (trace 3). Ca2+-dependent changes in tryptophan fluorescence emission are biphasic (see main text for details).

Protocol 7: Tryptophan Fluorescence of GCAPs

1. Dissolve a purified sample of GCAP in a buffer containing 50 mM HepesKOH, pH 7.4, 100 mM NaCl, and 1 mM DTT at a concentration of 2 µM. 2. Set the excitation wavelength of the fluorescence spectrometer to 280 nm. 3. Record first the emission spectrum between 300 and 450 nm of the buffer without GCAP. Then record the spectrum of the GCAP solution. 4. Correct the emission spectrum of GCAP by subtraction of the spectrum obtained with the buffer solution. This step will separate the tryptophan fluorescence signal from the Raman scattering signal of water that can interfere with the true fluorescence emission signal when low protein concentrations are used. 5. Vary the [Ca2+] in the medium by the use of a Ca2+/EGTA buffer system as described in Protocol 1. Repeat for every [Ca2+] the foregoing steps. Plot the maximum fluorescence intensity as a function of the [Ca2+], and determine the EC50 value.

A Ca2+ titration of GCAP2 is seen in figure 6.5. The slight increase at higher [Ca2+] is a typical observation made with GCAP1 (39,63,68). It was recently reported that binding of Ca2+ to EF-hand 4 in GCAP1 causes the structural movement around Trp94, which can be detected as the increase in fluorescence emission at [Ca2+] > 10 −6 M (39). Limited proteolysis has been used to investigate the accessibility of GCAPs for trypsin. Ca2+-bound and Ca2+-free wild-type GCAPs have different accessibilities for trypsin (68,80). Removing Ca2+ induces a conformational change that opens the interior of the protein and facilitates access of the protease. The comparison of

7315.indb 134

11/15/07 1:40:08 PM


135

wild-type and mutant GCAP forms allows conclusions about conformational stability and accessibility of certain regions in particular GCAPs.

6.3.3 Chemical Modification of GCAPs GCAPs contain 3–4 cysteines in their amino acid sequence. Therefore, chemical modification of cysteines opens several routes to investigating the molecular properties of these proteins. For example, introduction of spin labels and fluorescent and nonfluorescent dyes have been used to study Ca2+-dependent conformational changes associated with GCAP1 and GCAP2 (65,66). Cysteines in GCAPs exhibit different accessibilities to the thiol-modifying reagent 5,5′-dithiobis(2-nitrobenzoic acid) (DTNB) as it was demonstrated by the use of cysteine mutants of GCAP1. DTNB contains a disulfide group that undergoes a disulfide exchange reaction with free cysteines in a protein. This reaction produces the chromogenic substance 5-thio-2nitrobenzoic acid (TNB) and can be monitored by a change in absorbance at 412 nm. Modification of each thiol group in a protein produces exactly one molecule of chromogenic TNB from every DTNB and allows easy quantitation of the modification. Time-based absorbance measurements at 412 nm also enable kinetic measurements (66). For example, the four cysteines in GCAP1 react with different velocities with DTNB at different Ca2+ concentrations. Thus, they display different Ca2+ sensitivities to the molecular environment within the polypeptide chain.66 Protocol 8: Modification of GCAPs by DTNB

1. Prepare a fresh DTNB solution of 12 mM DTNB in 0.1 M Tris-HCl, pH 8.0. Sonicate the solution for several seconds at 80–100 W. 2. Prepare a solution of 2–3 µM GCAP in 50 mM Hepes, pH 7.4 and 100 mM NaCl. Add either 10 µM CaCl2 or 10 µM EGTA (calculate for a final volume of 2 mL in the cuvette, prepare first 1.99 mL, and degas the solution). 3. Stir the solution in the cuvette by a magnetic stirrer, and start the reaction by addition of 10 µL DTNB stock solution (60 µM final). Record the absorbance change at 412 nm for a sufficient time (5–10 min at least). Add 100 µM CaCl2 to an EGTA-containing cuvette or vice versa, and record change of absorption. 4. Calculate the TNB concentration using the molar absorption coefficient for TNB of ε412 = 13,600 M−1 × cm−1 and relate to the amount of modified cysteines.

An example of thiol modification of a GCAP1 mutant is shown in figure 6.6. The GCAP1 mutant ACAA contains only one of the four cysteines present in wild-type GCAP1. This cysteine is located in the nonfunctional EF-hand 1 and is easily accessible by DTNB. A Ca2+ titration experiment as shown in figure 6.6 reveals that the cysteine becomes blocked at higher [Ca2+] (> 100 µM), which mirrors a low affinity for Ca2+ binding of EF-hand 1.

7315.indb 135

11/15/07 1:40:09 PM

136

Signal Transduction in the Retina 0.05

ACAA

0.04

10 nM 2.8 mM

A412

0.03 0.02 0.01 0.00 0

2

4 6 Time (min)

8

10

12

Figure 6.6 Thiol reactivity of the cysteine residue at position 29 in the GCAP1 mutant ACAA. Three of the four cysteines in GCAP1 were substituted by alanine; the single cysteine at position 29 is accessible to DTNB as a function of free [Ca2+]. Recordings were started by the addition of DTNB (60 µM) to 2 µM ACAA. Free [Ca2+] was varied from 10 nM to 2.8 mM as indicated.

6.3.4 Reconstitution of Purified GCAPs with Membrane-Bound ROS-GCs The determination of guanylate cyclase activities to test modulating effects of GCAPs is performed by preparing membranes containing ROS-GCs and mixing the membranes with the corresponding GCAP solution. Sources of ROS-GC are either purified rod outer segments or cell membranes of HEK293, tsA or COS cells that contain heterologously expressed GCs. Purified samples of native ROS-GC1 in detergent solutions are not activated by GCAPs or GCAP-containing fractions (36). When ROS-GC1 samples were reconstituted in phospholipid membranes, addition of GCAPs also did not result in a Ca2+-dependent activation (Koch, unpublished observation). So far it is unclear why ROS-GC preparations fail to become activated by GCAPs when they are extracted by detergents or purified to homogeneity. Loss of a photoreceptor-specific factor is unlikely, because heterologously expressed ROS-GCs in HEK293, tsA, or COS cells are specifically activated by GCAPs and exhibit high activation rates at low [Ca2+] (40,46,57,59,68,74). It is conceivable that detergent solutions, which are necessary to extract the membrane-bound GCs, disturb a certain conformation that is necessary to interact with GCAPs or that they simply have a shielding effect by binding to the GCAP target site. Interestingly, soluble constructs of the cytoplasmic domain of ROS-GC1 interact with GCAP2 (62) and with the Ca2+-binding protein S100β (81) by similar affinities as observed for whole ROS-GC1 in membrane preparations. These constructs are also activated by GCAP2 at low [Ca2+] and by S100β at high [Ca2+]. However, the basal GC activities observed with soluble ROS-GC1 constructs were approximately tenfold lower than the activities of the whole enzyme. It was recently proposed that illuminated

7315.indb 136

11/15/07 1:40:10 PM


137

rhodopsin initiates an ATP-dependent preincubation phase during which GCs in ROS membranes are “primed.” This priming effect of ATP does not involve a phosphorylation reaction and results in a significantly enhanced stimulation of GCAPdependent ROS-GC activation (76–78). It remains to be shown whether priming of ROS-GCs is necessary for them to become activated by GCAPs and whether disruption of the membrane structure leads to a loss of “ATP priming.” Protocol 9: Reconstitution of GCAPs with ROS-GC Use of ROS in subsequent reconstitution studies with GCAPs requires the removal of native endogenous GCAPs from the ROS suspension. Two washing steps under low-salt conditions are usually sufficient to remove most of the endogenous GCAPs. One or two more washing steps might be additionally necessary when the remaining guanylate cyclase activity in ROS membranes exhibits a rather high Ca2+ sensitivity (more than twofold activation of cyclase). However, repeated washing can result in a gradual loss of GCAP sensitivity of ROS-GCs.

1. Prepare a sample of membranes containing heterologously expressed ROS-GC from tsA cells according to protocol 4. If you use HEK293 cells, perform a similar procedure. 2. If you use ROS membranes, take an aliquot of 500 µL purified ROS (6–10 mg/mL rhodopsin). Work under dim red light. Dilute with 2 mL of a lowsalt buffer (10 mM Hepes-KOH, pH 7.5, 1 mM DTT). Homogenize and centrifuge for 10 min (≥ 100,000 × g). Resuspend the pellet in 2 mL of lowsalt buffer, and repeat the centrifugation step. Resuspend the pellet in 250 µL resuspension buffer (50 mM Hepes-KOH, pH 7.5, 500 mM KCl, 20 mM NaCl, and 1 mM DTT). 3. Add 10 µL of a GCAP solution to 10 µL of a membrane suspension that contains one or both of ROS-GCs (e.g., tsA or ROS membranes). Vary the GCAP concentration within a range from 0 to 20 µM, depending on the properties of the specific GCAP isoform or GCAP mutant. Preincubate with 10 µL of a Ca2+-EGTA buffer (see protocol 1) for at least 5 min. 4. When you plan to measure a Ca2+-dependent activation profile, choose a saturating GCAP concentration and vary the free [Ca2+]. 5. Start the incubation reaction by adding 20 µL of the GC-assay buffer (see protocol 1). Perform the incubation and analysis of data as described in protocol 1. When working with ROS membranes, perform all steps under very dim red light until you quench the reaction. When working with heterologously expressed GCs, perform all steps under room light.

GCAP1 and GCAP2 display different activation profiles (40), when the activity of ROS-GCs in ROS membranes is measured at different free [Ca2+]. GCAP1 activates ROS-GCs at higher [Ca2+] than GCAP2 does (compare figures 6.7a and 6.7b). In addition to this difference in Ca2+ sensitivity, both GCAPs differ in the influence of their myristoyl groups on their catalytic efficiency, their monomer–dimer equilibria and their target recognition sites in ROS-GC1 (40,59–61,67,68).

7315.indb 137

11/15/07 1:40:10 PM

138 GC Activity (nmol cGMP × min–1 × mg Rh–1)

Signal Transduction in the Retina 7 6 5

GCAP2

4 3 2 1 0 100

101

102

103

104

105

[Ca2+] (nM)

GC Activity (nmol cGMP × min–1 × mg Rh–1)

Figure 6.7a Reconstitution of ROS-GCs in rod outer segment membranes with 3 µM recombinant myristoylated GCAP2. GC activity was measured at the indicated free [Ca2+]. Activation was half maximal at 78 nM. 12 10

GCAP1-D6S

8 6 4 2 0 100

101

102

103 [Ca2+]

104

105

106

(nM)

Figure 6.7b Reconstitution of ROS-GCs in rod outer segment membranes with 3 µM myristoylated GCAP1-D6S. GC activity was measured at the indicated free [Ca2+]. Activation was half maximal at 855 nM.

ACKNOWLeDGMENTS We thank Doris Höppner-Heitmann (IBI-1, Forschungszentrum Jülich, Germany) and Werner Säftel (University of Oldenburg) for excellent technical assistance. Research in the laboratory of Karl-Wilhelm Koch was funded by several grants of the Deutsche Forschungsgemeinschaft (DFG), an INTAS grant, and a grant from the EWE-Stiftung. We also acknowledge support from the Forschunsgszentrum Jülich.

7315.indb 138

11/15/07 1:40:12 PM


139

REFERENCES

7315.indb 139

1. Sunahara, R.K. and Taussig, R., Isoforms of mammalian adenylyl cyclase: multiplicities of signaling, Mol Interventions, 2, 168, 2002. 2. Koesling, D., Studying the structure and regulation of soluble guanylyl cyclase, Methods, 19, 485, 1999. 3. Lucas, K.A., Pitari, G.M., Kazerounian, S., Ruiz-Stewart, I, Park, J., Schulz, S., Chepenik, K.P., and Waldman, S.A., Guanylyl cyclases and signaling by cyclic GMP, Pharmacol Rev, 52, 375, 2000. 4. Wedel, B.J. and Garbers, D.L., The guanylyl cyclase familiy at Y2K, Annu Rev Physiol, 63, 215, 2001. 5. Potter, L.R., Domain analysis of human transmembrane guantylyl cyclase receptors: implications for regulation, Front Biosci, 10, 1205, 2005. 6. Fitzpatrick, D.A., O’Halloran, D.M., and Burnell, A.M., Multiple lineage specific expansions within the guanylyl cyclase gene family, BMC Evolut Biol, 6, 1, 2006. 7. Pugh, E.N., Jr., Duda, T., Sitaramayya, A., and Sharma, R.K., Photoreceptor guanylate cyclases: a review, Biosci Rep, 17, 429, 1997. 8. Koch, K.-W., Duda, T., and Sharma, R.K., Photoreceptor specific guanylate cyclases in vertebrate phototransduction, Mol Cell Biochem, 230, 97, 2002. 9. Sharma, R.K., Duda, T., Venkataraman, V., and Koch, K., Calcium-modulated mammalian membrane guanylate cyclase ROS-GC transduction machinery in sensory neurons: a universal concept, Curr Top Biochem Res, 6, 111, 2004. 10. Francis, S.H. and Corbin, J.D., Structure and function of cyclic nucleotide-dependent protein kinases, Annu Rev Physiol, 56, 237, 1994. 11. Lincoln, T.M., Dey, N., and Sellak, H., Signal transduction in smooth muscle. Invited Review: cGMP-dependent protein kinase signaling mechanisms in smooth muscle: from the regulation of tone to gene expression, J Appl Physiol, 91, 1421, 2001. 12. Kaupp, U.B. and Seifert, R., Cyclic nucleotide-gated ion channels, Physiol Rev, 82, 769, 2002. 13. Kramer, R.H. and Molokanova, E., Modulation of cyclic nucleotide-gated channels and regulation of vertebrate phototransduction, J Exp Biol, 204, 2921, 2001. 14. Sonderling, S.H. and Beavo, J.A., Regulation of cAMP and cGMP signaling: new phosphodiesterases and new functions, Curr Opin Cell Biol, 12, 174, 2000. 15. Rybalkin, S.D., Yan, C., Bornfeldt, K.E., and Beavo, J.A., Cyclic GMP phosphodiesterases and regulation of smooth muscle function, Circ Res, 93, 280, 2003. 16. de Rooij, J., Zwartkruis, F.J.T., Verheijen, M.H.G., Cool, R.H., Nijman, S.M.B., Wittinghofer, A., and Bos, J.L., Epac is a Rap1 guanine-nucleotide-exchange factor directly by cyclic AMP, Nature, 396, 474, 1998. 17. Margulis, A., Sharma, R.K., and Sitaramayya, A., Nitroprusside-sensitive and intensive guanylate cyclases in retinal rod outer segments, Biochem Biophys Res Commun, 185, 909, 1992. 18. Ahmad, I. and Barnstable, C.J., Differential laminar expression of particulate and soluble guanylate cyclase genes in rat retina, Exp Eye Res, 56, 51, 1993. 19. Koch, K.-W., Lambrecht, H.-G., Haberecht, M., Redburn, D., and Schmidt, H.H.H.W., Functional coupling of a Ca2+/calmodulin-dependent nitric oxide synthase and a soluble guanylyl cyclase in vertebrate photoreceptor cells, EMBO J, 13, 3312, 1994. 20. Haberecht, M.F., Schmidt, H.H.H.W., Mills, S.L., Massey, S.C., Nakane, M., and Redburn-Johnson, D.A., Localization of nitric oxide synthase, NADPH diaphorase and soluble guanylyl cyclase in adult rabbit retina, Vis Neurosci, 15, 881, 1998. 21. Gotzes, S., de Vente, J., and Müller, F., Nitric oxide modulates cGMP levels in neurons of the inner and outer retina in opposite ways, Vis Neurosci, 15, 945, 1998.

11/15/07 1:40:12 PM

140


22. Donovan, M., Carmody, R.J., and Cotter, T.G., Light-induced photoreceptor apoptosis in vivo requires neuronal nitric-oxide synthase and guanylate cyclase activity and is caspase-3-independent, J Biol Chem, 276, 23000, 2001. 23. Olshevskaya, E.V., Ermilov, A.N., and Dizhoor, A.M., Factors that affect regulation of cGMP synthesis in vertebrate photoreceptors and their genetic link to human retinal degeneration, Mol Cell Biochem, 230, 139, 2002. 24. Palczewski, K., Sokal, I., and Baehr, W., Guanylate cyclase-activating proteins: structure, function, and diversity, Biochem Biophys Res, 322, 1123, 2004. 25. Koch, K.-W., GCAPs, the classical neuronal calcium sensors in the retina, Calcium Binding Proteins, 1, 3, 2006. 26. Rollin, R., Mediero, A., Roldán-Pallarés, M., Fernández-Cruz, A., and Fernández-Durango, R., Natriuretic peptide system in the human retina, Mol Vision, 10, 15, 2004. 27. Yu, Y.-Ch., Cao, L.-H., and Yang, X.-L., Modulation by brain natriuretic peptide of GABA receptors on rat retinal ON-type bipolar cells, J Neurosci, 26, 696, 2006. 28. Pugh, E.N., Jr. and Lamb, T.D., Phototransduction in vertebrate rods and cones: molecular mechanisms of amplification, recovery and light adaption, Handbook of Biological Physics, Vol. 3 (Eds. Stavenga, DeGrip, and Pugh Jr., Elsevier Science B.V.) 2000, chap. 5. 29. Burns, M.E. and Baylor, D.A., Activation, deactivation, and adaptation in vertebrate photoreceptor cells, Annu Rev Neurosci, 24, 779, 2001. 30. Fleischmann, D. and Denisevich, M., Guanylate cyclase of isolated bovine retinal rod axonemes, Biochemistry, 18, 5060, 1979. 31. Koch, K.-W. and Stryer, L., Highly cooperative feedback control of retinal rod guanylate cyclase by calcium ions, Nature, 334, 64, 1988. 32. Gorczyca, W.A., van Hooser, J.P., and Palczewski, K., Nucleotide inhibitors and activators of retinal guanylyl cyclase, Biochemistry, 33, 3217, 1994. 33. Koch, K.-W., Eckstein, F., and Stryer L., Stereochemical course of the reaction catalyzed by guanylate cyclase from bovine retinal rod outer segments, J Biol Chem, 265, 9659, 1990. 34. Wolbring, G. and Schnetkamp, P.P.M., Activation by PKC of the Ca2+-sensitive guanylyl cyclase in bovine retinal rod outer segments measured with an optical assay, Biochemistry, 34, 4689, 1995. 35. Sharma, R.K., Marala, R.B., and Duda, T.M., Purification and characterization of the 180-kDa membrane guantylate cyclase containing atrial natriuretic factor receptor from rat adrenal gland and its regulation by protein kinase C, Steroids, 53, 437, 1989. 36. Koch, K.-W., Purification and identification of photoreceptor guanylate cyclase, J Biol Chem, 266, 8634, 1991. 37. Tsien, R. and Pozzan, T., Measurement of cytosolic free Ca2+ with Quin2, Methods Enzymol, 172, 230, 1989. 38. Lambrecht, H.-G. and Koch, K.-W., A 26 kd calcium binding protein from bovine rod outer segments as modulator of photoreceptor guanylate cyclase, EMBO J, 10, 793, 1991. 39. Peshenko, I.V. and Dizhoor, A.M., Ca2+-and Mg2+-binding properties of GCAP-1: evidence that Mg2+-bound form is the physiological activator of photoreceptor guanylyl cyclase, J Biol Chem, 281, 23830, 2006. 40. Hwang, J.-Y., Lange, C., Helten, A., Höppner-Heitmann, D., Duda, T., Sharma, R.K., and Koch, K.-W., Regulatory modes of rod outer segment membrane guanylate cyclase differ in catalytic efficiency and Ca2+-sensitivity, Eur J Biochem, 270, 3814, 2003. 41. Hayashi, F. and Yamazaki, A., Polymorphism in purified guanylate cyclase from vertebrate rod photoreceptors, Biochemistry, 88, 4746, 1991. 42. Margulis, A., Goraczniak, R.M., Duda, T., Sharma, R.K., and Sitaramayya, A., Structural and biochemical identity of retinal rod outer segment membrane guanylate cyclase, Biochem Biophys Res Commun, 194, 855, 1993.

7315.indb 140

11/15/07 1:40:12 PM


141

43. Koch, K.-W., Stecher, P., and Kellner, R., Bovine retinal rod guanyl cyclase represents a new N-glycosylated subtype of membrane-bound guanyl cyclases, Eur J Biochem, 222, 589, 1994. 44. Shyjan, A.W., de Sauvage, F.J., Gillett, N.A., Goeddel, D.V., and Lowe, D.G., Molecular cloning of a retina-specific membrane guanylyl cyclase, Neuron, 9, 727, 1992. 45. Goraczniak, R.M., Duda, T., Sitaramayya, A., and Sharma, R.K., Biochem J, 302, 455, 1994. 46. Lowe, D.G., Dizhoor, A.M., Liu, K., Gu, Q., Spencer, M., Laura, R., Lu, L., and Hurley, J.B., Proc Natl Acad Sci U.S.A., 92, 5535, 1995. 47. Yang, R.-B., Foster, D.C., Garbers, D.L., and Fülle, H.-J., Two membrane forms of guanylyl cyclase found in the eye, Proc Natl Acad Sci U.S.A., 92, 602, 1995. 48. Goraczniak, R., Duda, T., and Sharma, R.K., Structural and functional characterization of a second subfamily member of the calcium-modulated bovine rod outer segment membrane guanylate cyclase, ROS-GC2, Biochem Biophys Res Commun, 234, 666, 1997. 49. Yang, R.-B. and Garbers, D.L., Two eye guanylyl cyclases are expressed in the same photoreceptor cells and form homomers in preference to heteromers, J Biol Chem, 272, 13738, 1997. 50. Chen, C. and Okayama, H., High-efficiency transformation of mammalian cells by plasmid DNA, Mol Cell Biol, 7, 2745, 1987. 51. Senin, I.I., Koch, K.-W., Akhtar, M., and Philippov, P.P., Ca2+-dependent control of rhodopsin phoshorylation: recoverin and rhodopsin kinase, Adv Exp Med Biol, 514, 69, 2002. 52. Hsu, Y.-T. and Molday, R.S., Modulation of the cGMP-gated channel of rod photoreceptor cells by calmodulin, Nature, 361, 76, 1993. 53. Weitz, D., Zoche, M., Müller, F., Beyermann, M., Körschen, H.-G., Kaupp, U.B., and Koch, K.-W., Calmodulin controls the rod photoreceptor CNG channel through an unconventional binding site in the N-terminus on the ß-subunit, EMBO J, 17, 2273, 1998. 54. Grunwald, M.E., Yu, W.-P., Yu, H.-H., and Yau, K.-W., Identification of a domain on the ß-subunit of the rod cGMP-gated cation channel that mediates inhibition by calciumcalmodulin, J Biol Chem, 273, 9148, 1998. 55. Imanishi, Y., Yang, L., Sokal, I., Filipek, S., Palczewski, K., and Baehr, W., Diversity of guanylate cyclase-activating proteins (GCAPs) in teleost fish: charcterization of three novel GCAPs (GCAP4, GCAP5, GCAP7) from zebrafish (Danio rerio) and prediction of eight GCAPs (GCAP1-8) in pufferfish (Fugu rubripes), J Mol Evol, 59, 204, 2004. 56. Gorczyca, W.A., Gray-Keller, M.P., Detwiler, P.B., and Palczewski, K., Purification and physiological evaluation of a guanylate cyclase activating protein from retinal rods, Proc Natl Acad Sci U.S.A., 91, 4014, 1994. 57. Dizhoor, A.M., Lowe, D.G., Olshevskaya, E.V., Laura, R.P., and Hurley, J.B., The human photoreceptor membrane guanylyl cyclase, RetGC, is present in outer segments and is regulated by calcium and a soluble activator, Neuron, 12, 1345, 1994. 58. Frins, S., Bönigk, W., Müller, F., Kellner, R., and Koch, K.-W., Functional characterization of a guanylyl cyclase-activating protein from vertebrate rods, J Biol Chem, 271, 8022, 1996. 59. Lange, C., Duda, T., Beyermann, M., Sharma, R.K., and Koch, K.-W., Regions in vertebrate photoreceptor guanylyl cyclase ROS-GC1 involved in CA2+-dependent regulation by guanylyl cyclase-activating protein GCAP-1, FEBS Lett, 460, 27, 1999. 60. Sokal, I., Haeseleer, F., Arendt, A., Adman, E.T., Hargrave, P.A., and Palczewski, K., Identification of a guanylyl cyclase-activating protein-binding site within the catalytic domain of retinal guanylyl cyclase 1, Biochemistry, 38, 1387, 1999. 61. Krylov, D.M. and Hurley, J.B., Identification of proximate regions in a complex of retinal guanyly cyclase 1 and guanylyl cyclase-activating protein-1 by a novel mass spectrometry-based method, J Biol Chem, 276, 30648, 2001.

7315.indb 141

11/15/07 1:40:13 PM

142


62. Duda, T., Fik-Rymarkiewicz, E., Venkataraman, V., Kishnan, R., Koch, K.-W., and Sharma, R.K., The calcium-sensor guanylate cyclase activating protein type 2 specific site in rod outer segment membrane guanylate cyclase type 1, Biochemistry, 44, 7336, 2005. 63. Otto-Bruc, A., Buczylko, J., Surgucheva, I., Subbaraya, I., Rudnicka-Nawrot, M., Crabb, J.W., Arendt, A., Hargrave, P.A., Baehr, W., and Palczewski, K., Functional reconstitution of photoreceptor guanylate cyclase with native and mutant forms of guanylate cyclase-activating protein 1, Biochemistry, 36, 4295, 1997. 64. Sokal, I., Otto-Bruc, A.E., Surgucheva, I., Verlinde, C.L.M., Wang, C.-K., Baehr, W., and Palczewski, K., Conformational changes in guanylyl cyclase-activating protein 1 (GCAP1) and its tryptophan mutants as a function of calcium concentration, J Biol Chem, 274, 19829, 1999. 65. Sokal, I., Li, N., Klug, C.S., Filipek, S.B., Hubbell, W.L., Baehr, W., and Palczewski, K., Calcium-sensitive regions of GCAP1 as observed by chemical modifications, fluorescence, and EPR spectroscopies, J Biol Chem, 276, 43361, 2001. 66. Hwang, J.-Y., Schlesinger, R., and Koch, K.-W., Calcium-dependent cysteine reactivities in the neuronal calcium sensor guanylate cyclase-activating protein 1, FEBS Lett, 508, 355, 2001. 67. Ermilov, A.N., Olshevskaya, E.V., and Dizhoor, A.M., Instead of binding calcium, one of the EF-hand structures in guanylyl cyclase activating protein-2 is required for targeting photoreceptor guanylyl cyclase, J Biol Chem, 276, 48143, 2001. 68. Hwang, J.-Y, Schlesinger, R., and Koch, K.-W., Irregular dimerization of guanylate cyclase-activating protein 1 mutants causes loss of target activation, Eur J Biochem, 271, 3785, 2004. 69. Duda, T. and Koch, K.-W., Retinal diseases linked with photoreceptor guanylate cyclase, Mol Cell Biochem, 230, 129, 2002. 70. Mendez, A., Burns, M.E., Sokal, I., Dizhoor, A.M., Baehr, W., Palczewski, K., Baylor, D.A., and Chen, J., Role of guanylate cyclase-activating proteins (GCAPs) in setting the flash sensitivity of rod photoreceptors, Proc Natl Acad Sci U.S.A., 98, 9948, 2001. 71. Howes, K.A., Pennesi, M.E., Sokal, I., Church-Kopish, J., Schmidt, B., Margolis, D., Frederick, J.M., Rieke, F., Palczewski, K., Wu, S.M., Detwiler, P.B., and Baehr, W., GCAP1 rescues rod photoreceptor response in GCAP1/GCAP2 knockout mice, EMBO J, 21, 1545, 2002. 72. Burns, M.E., Mendez, A., Chen, J., and Baylor, D.A., Dynamics of cyclic GMP synthesis in retinal rods, Neuron, 36, 81, 2002. 73. Pennesi, M.E., Howes, K.A., Baehr, W., and Wu, S.M., Guanylate cyclase-activating protein (GCAP) 1 rescues cone recovery kinetics in GCAP1/GCAP2 knockout mice, Proc Natl Acad Sci U.S.A., 100, 6783, 2003. 74. Duda, T., Goraczniak, R., Surgucheva, I., Rudnicka-Nawrot, M., Gorczyca, W.A., Palczewski, K., Sitaramayya, A., Baehr, W., and Sharma, R.K., Calcium modulation of bovine photoreceptor guanylate cyclase, Biochemistry, 35, 8478, 1996. 75. Schrem, A., Lange, C., Beyermann, M., and Koch K.-W., Identification of a domain in guanylyl cyclase-activating protein 1 that interacts with a complex of guanylyl cyclase and tubulin in photoreceptors, J Biol Chem, 274, 6244, 1999. 76. Yamazaki, A., Yu, H., Yamazaki, M., Honkawa, H., Matsuura, I., Usukura, J., and Yamazaki, R.K., A critical role for ATP in the stimulation of retinal guanylyl cyclase by guanylyl cyclase-activating proteins, J Biol Chem, 278, 33150, 2003. 77. Yamazaki, M., Usukura, J., Yamazaki, R.K., and Yamazaki, A., ATP binding is required for physiological activation of retinal guanylate cyclase, Biochem Biophys Res Commun, 338, 1291, 2005. 78. Yamazaki, A., Yamazaki, M., Yamazaki, R.K., and Usukura, J., Illuminated rhodopsin is required for strong activating of retinal guanylate cyclase by guanylate cyclaseactivating proteins, Biochemistry, 45, 1899, 2006.

7315.indb 142

11/15/07 1:40:13 PM


143

79. Hwang, J.-Y. and Koch, K.-W., Calcium- and myristoyl-dependent properties of guanylate cyclase-activating protein-1 and protein-2, Biochemistry, 41, 13021, 2002. 80. Rudnicka-Nawrot, M., Surgucheva, I., Hulmes, J.D., Haeseleer, F., Sokal, I., Crabb, J.W., Baehr, W., and Palczewski, K., Changes in biological activity and folding of guanylate cyclase-activating protein 1 as a function of calcium, Biochemistry, 37, 248, 1998. 81. Duda, T., Koch, K.-W., Venkataraman, V., Lange, C., Beyermann, M., and Sharma, R.K., Ca2+ sensor S100β-modulated sites of membrane guanylate cyclase in the photoreceptor-bipolar synapse, EMBO J, 21, 2547, 2002.

7315.indb 143

11/15/07 1:40:13 PM

7315.indb 144

11/15/07 1:40:13 PM

7

Transgenic Strategies for Analysis of Photoreceptor Function Janis Lem and Kibibi Rwayitare

Contents 7.1

Introduction.................................................................................................. 145 7.1.1 Background and Aims...................................................................... 145 7.1.2 Overview of Methodologies.............................................................. 147 7.1.2.1 Pronuclear Microinjection................................................... 147 7.1.2.2 Gene Targeting in ES Cells (Followed by Injection into Blastocysts)................................................................... 148 7.2 Experimental Strategies............................................................................... 150 7.2.1 Knockout Mice: Ablation of Endogenous Protein Expression......... 150 7.2.2 Transgene Knockdown of Endogenous Protein (Hypomorphs)....... 151 7.2.3 Knockin Mutations........................................................................... 153 7.2.4 Overexpression of WT or Mutant Protein........................................ 153 7.2.4.1 Promoter Analysis............................................................... 153 7.2.4.2 Dominant Mutants............................................................... 154 7.2.4.3 Wild-Type Gene Expression................................................ 154 7.2.4.4 Loss-of-Function Mutations................................................ 155 7.2.5 Conditional or Regulated Transgene Expression.............................. 155 7.3 Conclusions.................................................................................................. 156 Acknowledgments................................................................................................... 157 References............................................................................................................... 157

7.1 Introduction 7.1.1

Background and Aims

The recently completed sequencing of the human and mouse genomes has propelled research to focus on the study of gene function in normal physiology and disease processes. Transgenic mice have an extensive track record as powerful 145

7315.indb 145

11/15/07 1:40:26 PM

146


tools for elucidating gene structure and function. Early transgenic mouse studies focused on genetic dissection of gene structure, with an emphasis on identifying genetic elements of genes that regulated cell- and tissue-specific expression. Over the past two decades, with the discovery of novel genes, the focus of transgenic mouse research has increasingly shifted toward examination of gene function. Technological advances currently allow direct manipulation of the endogenous gene to examine pathological effects in vivo. Transgenic mouse models allow an integrated approach to the study of human disease processes. Mouse models permit detailed characterization of changes in tissue morphology, biochemical pathways, and physiologic changes related to the disease state. These can be determined along a developmental time course, potentially leading to the development of diagnostic biomarkers that can define early or late stages of disease. Identifying the earliest cellular changes can help define the molecular basis of genetic disease, distinguishing secondary events that may be an effect rather than a cause of disease. Understanding the molecular mechanisms can lead to the design of rational therapies. Finally, transgenic mouse models provide a resource to test new therapies for effective treatment of human diseases. For several reasons, the mouse is the most common transgenic animal model produced. Compared to large mammalian animal models (e.g., monkeys, sheep, and cows), mice are relatively cheap to purchase and house. Although mouse models do not always accurately model human disease, more often than not they share conserved biochemical pathways that reveal molecular mechanisms shared with other vertebrates. The most powerful reason to use mice is the well-characterized genetics of the mouse. Inbred mouse lines eliminate genetic background variations that confound interpretation of genetic inheritance in human disease and in most other mammalian models. Having transgenic mice on a defined genetic background is key to the identification of modifier genes that can influence disease susceptibility. Furthermore, a uniform genetic background is essential for definitive analysis of the contribution of epigenetic factors to disease pathology. The term transgene denotes the genetic material that is introduced into an organism. The two major methods used for the introduction of genetic material into mice are by pronuclear microinjection or by gene targeting in embryonic stem (ES) cells, followed by injection of the targeted ES cells into blastocysts. The major difference between the two methods is that in pronuclear microinjections, genetic material randomly integrates into the mouse genome in an additive manner. In contrast, gene targeting uses homologous recombination to substitute genetic material nonrandomly into the mouse genome. In this chapter, we present an overview of transgenic strategies for the analysis of gene function. We will discuss several commonly used transgenic strategies for analyzing gene function, including total ablation of endogenous protein expression, reduced levels of endogenous protein expression, introduction of a mutation in the endogenous gene, overexpression of normal or mutant protein, and controlled, regulated expression of transgene expression. Each of the methodologies has strengths and limitations that must be considered in the design and interpretation of a transgenic experiment. This chapter does not provide experimental details on ES cell culture methods (1,2) or detailed technical methodologies for the preparation of embryos for microinjection,

7315.indb 146

11/15/07 1:40:27 PM

Transgenic Strategies for Analysis of Photoreceptor Function

147

as those are easily found in manuals on transgenic methodology (1,3,4). Instead, this review focuses on considerations for selecting an experimental strategy.

7.1.2 Overview of Methodologies 7.1.2.1 Pronuclear Microinjection In pronuclear microinjections, linearized transgene DNA is microinjected into fertilized single-cell wild-type mouse embryos (figure 7.1). Characteristic of pronuclear microinjections is the random integration of transgene DNA into the mouse genome. The integrated transgene is subject to regulation by DNA surrounding the site of insertion. Consequently, the transgene is not completely regulated in a spatial or temporal manner that models that of the endogenous gene. Transgenic mice produced in this manner can show substantial variation in spatial expression patterns and the level of transgene expression. This property can be used to advantage by correlating the level of expression with severity of phenotype. Pronuclear injection methodology lends itself to certain types of studies, such as overexpression of wild-type or dominant mutant genes (discussed in the following text). However, it is not possible to study recessive mutations on a wild-type genetic background by using this methodology. For pronuclear microinjections, the transgene comprises three parts: an upstream regulatory promoter, the structural gene, and the polyadenylation signal. The structural gene may derive from genomic DNA, complementary DNA (cDNA), or a chimeric fusion of two genes. However, it should be noted that the presence of introns are Day 1

Donor females are hormone primed and mated with stud males.

Day 5

Two-cell embryos are transferred back into the oviducts of 0.5d pseudopregnant female mice. Nineteen days later, pups are born.

Day 3

Injected single cell embryos are cultured overnight to the two-cell stage, identifying viable embryos that survived microinjection.

At 0.5d p.c, embryos are harvested at the one-cell stage. Day 4

The transgene DNA is introduced through a fine capillary needle into the larger pronucleus, typically the male pronucleus.

Day 25

Figure 7.1 (See accompanying color CD.) Schematic of a pronuclear microinjection. On day one, 3- to 4-weeks-old donor females are injected with pregnant mare’s serum. On day three, females receive a dose of human chorionic gonadotropin and are mated with stud males. The following day, embryos at the one-cell stage are harvested for injection. Purified linearized transgene DNA is injected into the male pronucleus of the embryo. After injection, the embryos are incubated in a 5% CO2 incubator overnight. Embryos that develop to the twocell stage embryos are transferred into 0.5 day postcoitus pseudopregnant females. Potential founders are born 19 days later.

7315.indb 147

11/15/07 1:40:27 PM

148


reported to increase transcriptional efficiency (5). There are few restrictions on species specificity for gene expression (i.e., genes of mouse, rat, human, bovine, etc., origin may be used for microinjection into mouse embryos). The promoter defines the tissue and cell specificity as well as transgene expression level. If possible, it is desirable to use a promoter that has previously been tested in an animal model for cell specificity and promoter strength. However, regulatory elements that define temporal and spatial expression may be as far as 50 kb upstream (6,7). The unintentional deletion of an upstream regulatory element can result in transgene expression in a subset of cells that normally express the gene of interest, or show early or delayed temporal expression. The effects of these changes can also contribute to the observed phenotype. Expression of the transgene should be tested in a cell culture system prior to microinjection whenever possible, especially when producing chimeric fusion genes, to avoid potential issues such as artificial start or splice sites or changes in translation frame that are inadvertently introduced during the cloning process. Such changes could prevent expression of a functional transgene protein. 7.1.2.2 Gene Targeting in ES Cells (Followed by Injection into Blastocysts) Homologous recombination directly targets the endogenous gene and substitutes existing genetic material. Gene targeting may involve the introduction of a null mutation to produce a “knockout” transgenic mouse (8–17) (that lacks functional protein), the insertion of a dominant or recessive mutation within the endogenous gene to produce a “knockin” transgenic mouse (18), or the replacement of a mutant gene with a normal wild-type gene to correct a genetic defect (19). In homologous recombinants, the transgene is regulated under its normal endogenous promoter in a manner that is temporally and spatially consistent with the wild-type endogenous gene. However, the production of homologously recombinant transgenic mice is a labor-intensive process. First, ES cells with the desired homologous recombination event must be produced (1,2,20). Most embryonic stem cells are derived from one of several 129SV mouse substrains. ES cells from other mouse strains, including C57BL/6, DBA-1, FVB/n, Balb/C, 129 × C57Bl/6, and 129 × Balb/C, are also commercially available (table 7.1). ES cells are fastidious and must be meticulously cultured to retain their pluripotency. With continued passage, all ES cell lines become aneuploid and will eventually lose their ability for germ-line transmission. It is important to know for your particular ES cell line the recommended passage numbers that will retain germ-line competence. Using ES cells beyond the recommended passage number will decrease the likelihood of successful germ-line transmission. After ES cell clones with the desired homologous recombination event are identified, they are injected into the cavity of blastocyst stage embryos (figure 7.2). The injected embryos are surgically reimplanted into the uterus of foster mothers and allowed to develop to term (1). Offspring that have successfully taken up the ES cells produce chimeric founder mice. Additional rounds of breeding are required to obtain mice heterozygous or homozygous for the transgene mutation before animals are ready for phenotypic analysis. Breeding chimeric mice to homozygosity is a time-consuming process. Recently, Valenzuela and colleagues have shown that laser-assisted injection of ES cells into

7315.indb 148

11/15/07 1:40:28 PM


149

Table 7.1 ES Cell Lines and BAC Libraries Source

ES Cell Type

BAC Library

Specialty Media www.specialtymedia.com

C57BL/6 DBA-1 129SvEv W4 (129S6/SVEvTac)

Invitrogen, CHORIa Stratagene, CHORI

Taconic www.taconic.com ATCC http://stemcells.atcc.org

Thrombogenics www.thromb-x.com

Open Biosystems www.openbiosystems.com a

Day 1

ES-C57BL/6 J1 (129S4/SvJae) RW.4 (129X1/SvJ) 7ACS/EYFP (129X1x129S1) R1/E (129X1 x 129S1) Balb/C 129SvEv C57BL/6 FVB/N C57BL/6 129

Invitrogen

Stratagene Invitrogen

Invitrogen

bacpac.chori.org

Day 3 Donor females are hormone primed and mated with stud males.

Injected blastocysts are transferred back into the uterus of 2.5d pseudopregnant female mice. Seventeen days later, pups are born.

At 3.5d p.c., embryos are harvested at the blastocyst stage.

Day 7

Twelve to fifteen targeted ES cells are collected and introduced with the use of a tailored glass capillary needle onto the cell mass of an expanded blastocyst.

Day 24

Figure 7.2 (See accompanying color CD.) Schematic of a blastocyst injection. On day one, 3- to 4-week-old donor females are injected with pregnant mare’s serum. On day three, females receive a dose of human chorionic gonadotropin and are mated with stud males. On day seven, embryos at the blastocyst stage are harvested and injected with targeted embryonic stem cells. The injected blastocysts are transferred into the uterus of 2.5 day postcoitus pseudopregnant females. Seventeen days later, chimeric pups are born.

eight-cell embryos instead of blastocysts yielded transgenic mice that are almost wholly derived from ES cells. This permits phenotyping in the first generation of mice, because the transgene is essentially present in the heterozygous state. This represents a considerable savings in terms of time and effort (21).

7315.indb 149

11/15/07 1:40:29 PM

150


7.2 Experimental Strategies 7.2.1 Knockout Mice: Ablation of Endogenous Protein Expression Complete ablation of a protein is most often achieved by introduction of a null mutation within the endogenous gene. The null mutation is introduced by homologous recombination in embryonic stem cells (1). The frequency of homologous recombination is 1 in 106 to 107 cells. Thus, it is beneficial to optimize for successful homologous recombination in ES cells. The following conditions are recommended: The targeting construct should be made from mouse DNA originating from the same mouse strain as the ES cell line (22) (i.e., isogenic DNA). Bacterial artificial chromosome (BAC) libraries have been prepared from some 129Sv and C57Bl/6 mouse lines, facilitating rapid cloning of genes (23,24) (table 7.1). For homologous recombination to occur, flank the mutation with isogenic DNA with total homology between 5 and 10 kb in length (25,26) (figure 7.3). For a knockout targeting construct, a positive drug-selectable marker such as neomycin, hygromycin, or puromycin is inserted to replace exons known to be critical for gene function or to replace a region, including the translation start site. Inclusion of a positive drug-selectable marker increases the frequency of homologous recombination to 1 in 102 to 103 cells. Insertion of a negative selectable marker such as thymidine kinase or diphtheria toxin is reported to increase the frequency of homologous recombinants by approximately twofold, but also has a higher frequency of undesired rearrangements (26). Consideration must also be given as to whether a partially functional protein product might be produced instead of the desired null mutation. Targeting constructs are typically introduced into ES cells by electroporation (2,20). After 5 to 7 days of drug selection, clonal isolates are picked, and each clone Wild Type Gene 5'

1

2

3

4

5

Positive Drug Marker

4

5


4

5

3'

Targeting Construct 5'

1

Negative Marker

3'

Targeted Gene 5'

1

3'

Figure 7.3 Schematic of a knockout strategy by homologous recombination. Using isogenic DNA, functional domains of the wild-type gene known to be required for function are substituted with a positive drug-selectable marker such as neomycin, hygromycin, or puromycin. A negative selectable marker such as thymidine kinase or diphtheria toxin may be added to the 5′ or 3′ end of the targeting construct. Homology arms flanking the selectable marker should be 5 to 10 kb in length.

7315.indb 150

11/15/07 1:40:29 PM


151

tested for the desired homologous recombination event by PCR analysis and confirmed by Southern blot analysis. Clones with the desired homologous recombination event are then microinjected into blastocysts to produce chimeric mice (figure 7.2). If competent for germ-line transmission, first generation (G1) offspring derived from chimeric founders are heterozygous for the mutation and are bred to homozygosity. Several genes of the phototransduction cascade have been knocked out, including rod opsin (12,14), transducin α-subunit (8), cGMP phosphodiesterase gamma-subunit (27), rhodopsin kinase (11), and arrestin (9), to name but a few. Complete knockouts can also reveal a developmental role for a protein, leading to embryonic or perinatal lethality or developmental abnormalities that may complicate the interpretation of function in adult tissues. To study gene function in adult tissues without the complication of embryonic lethality, it may be desirable to manipulate expression in adult tissues using conditional transgene expression methods (see following text).

7.2.2 Transgene Knockdown of Endogenous Protein (Hypomorphs) Heterozygous mice with a null mutation in a single allele often have a reduced level of protein expression. For example, heterozygous rhodopsin null mutant mice have approximately 50% of wild-type levels of rhodopsin (14,28). However, the reduction is not always 50% of wild-type levels, as the remaining functional allele may undergo a compensatory upregulation of gene expression. For instance, transgenic mice heterozygous for the transducin α-subunit null mutation upregulate the remaining wild-type allele to near normal levels (8). RNA interference methods have been used successfully in transgenic mice (29,30), with germ-line transmission. In brief, expression of short hairpin RNAs (shRNA) from DNA vectors results in the production of double-stranded premicroRNAs. The endogenous Rnase III enzyme Dicer processes the double-stranded RNAs into ~22 nucleotide short interfering RNAs (siRNAs). The siRNAs produced upon cleavage by Dicer are incorporated into the RNA-inducing silencing complex (RISC). The siRNAs target the RNA to be cleaved by the Argonaute protein within the RISC complex. The shRNA DNA vectors have been successfully introduced into the germ line of mice by pronuclear injection, stable transfection into embryonic stem cells followed by injection into blastocysts, or by direct lentiviral infection of ES cells or infection of embryos (figure 7.4). These methods result in the random integration of the shRNA construct into the mouse genome. Each method has its advantages and limitations, as discussed in the following text. The DNA vectors may be stably introduced by direct pronuclear microinjection of DNA vector shRNAs into one-cell embryos or by lentiviral infection (31) of embryos. Two methods are commonly used for lentiviral infection of embryos. Both methods are highly efficient for the production of transgenic mice (~70 to 80% of offspring are transgenic). The first method involves the removal of the zona pellucida by enzymatic methods, followed by lentiviral transduction. This method is relatively simple technically and does not require the use of micromanipulators. However, removal of the zona pellucida is detrimental to the survival of embryos, with significantly fewer

7315.indb 151

11/15/07 1:40:30 PM

152


(a)

(b)

Embryonic Stem Cells

Lentiviral Infection

Electroporation

One-cell Embryos

Multicopy Insertion

Single Copy Insertion

7+

1

2

8

6

7

9 10 11 12

13

14

15 16 17 18 19

X

Y

wt

3

4

wt

5

7+

Lentiviral Injection into the Perivitelline Space

Denudation and Lentiviral Infection

7+

No chromosomal = Reproducible Expression Segregation from Animal to Animal

1

2

8

6

7

9 10 11 12

13

14

15 16 17 18 19

X

Y

1,7,16+

3

11+

4

5

7,11+ 16,1,7+

1+

Chromosomal = Variable Expression Segregation Levels in Offspring

Figure 7.4 Introduction of short-hairpin RNA (shRNA) DNA vectors into the germ line of mice. (a) shRNA DNA vectors can be introduced by electroporation or lentiviral transduction of embryonic stem (ES) cells. The majority of clonal ES cell lines will integrate the shRNA DNA vector as a single copy. Injected ES cells will yield founders containing a single integration site. (b) ShRNA DNA vectors can also be introduced by the removal of the zona pellucida (denudation) by enzymatic methods followed by lentiviral transduction or by direct microinjection into the perivitelline space surrounding the embryo. These methods will yield founders with multiple integration sites. Chromosomal segregation in subsequent generations will result in offspring with multiple integration sites, each with variable levels of transgene expression.

embryos surviving. The second method involves direct injection of lentivirus into the perivitelline space surrounding the embryo. This method avoids removal of the zona pellucida, but requires the use of micromanipulators and is technically more difficult. Lentiviral transduction results in multiple integration sites of the lentiviral transgene, thus giving the high percentage of transgenic founders. However, breeding of founder mice results in the segregation of chromosomes carrying the integrated transgenes that can continue over several generations of breeding. Second- to third-generation offspring can have variable levels of transgene expression, as each has a different integration site. This complicates interpretation of the data, as there is wide animalto-animal variability. Transduction of embryonic stem cells with lentivirus has the advantage that a single transgene is integrated into the ES cell. Clonal ES cell lines, each with a single integration site, are subsequently injected into blastocysts to produce chimeric mice. Several different clones may be injected to obtain mouse lines with different integration sites and different levels of expression. Although more labor intensive, this method has the advantage that each founder yields offspring with the same transgene integration site and reproducibility upon phenotypic analysis. A range of knockdowns is often produced that is dependent upon the specificity of the shRNA and its level of expression upon stable integration into the genome.

7315.indb 152

11/15/07 1:40:31 PM


153

Thus, it is useful to correlate the degree of knockdown with the severity of phenotype. Knockdown is seldom complete, although it is possible to knock down expression to 5% of WT expression levels. A potential problem with this approach is that highly expressed proteins can be difficult to knock down. A phenotype will only be seen if sufficient knockdown is induced. Thus, if 10 to 20% of WT levels are sufficient to maintain wild-type function, a phenotype may not be observed. There is much interest in using RNA interference methods for therapeutic treatment of retinal degeneration (32). Current studies support feasibility for this approach (33–36).

7.2.3 Knockin Mutations For some gene products, overexpression or underexpression of a wild-type gene results in a pathological phenotype. For instance, overexpression of wild-type rhodopsin (37) or reduced expression (19) results in retinal pathology. Thus, it is important to properly regulate the level of expression of such genes from its normal endogenous promoter to separate the effects of the mutation itself from effects arising from altered levels of transgene expression. The knockin mutation is introduced by homologous recombination, followed by deletion of the drug-selectable marker. A rhodopsin palmitoylation mutation was studied using this methodology (18). These animals revealed a subtle regulatory role for palmitoylation in phototransduction that most likely would have been masked had the mutation been introduced by pronuclear microinjection. Similarly, farnesylation of the gamma subunit of transducin was studied, revealing regulatory effects on light adaptation (38). Knockin mutations in several other photoreceptor genes support their role in retinal degenerative disease (39–44).

7.2.4 Overexpression of WT or Mutant Protein Information about mutant protein function can also be obtained by overexpressing either the wild-type protein or a dominant mutant of the protein. Embryo donor lines used for pronuclear microinjections should be checked for the presence of the recessive rd (retinal degeneration) allele (for example, the FVB/n mouse line) (45), as this could yield retinal degenerations unrelated to the transgene. The transgene is typically made using a promoter that has previously been characterized in regard to tissue-specific expression, level of expression (promoter strength), and developmental time course of expression. For promoters that have not been previously characterized, it is useful to utilize a reporter gene to define tissue specificity and timing of gene expression. 7.2.4.1 Promoter Analysis Pronuclear microinjection techniques lend themselves to the identification of regulatory elements within the promoter of genes. The earliest studies involved sequential truncation of portions of the 5′ upstream regulatory promoter region. Tissue or cell specificity of cloned promoters can be defined by microinjection of fusions of the promoter region with various reporter genes. Commonly used reporter genes are β-galactosidase, luciferase, and green fluorescent protein (GFP) derivatives. Alternatively, epitope tags such as the myc or hemagglutinin (HA) tags may be incorporated

7315.indb 153

11/15/07 1:40:31 PM

154


into the structural gene. Not only do the reporter genes track cell- and tissue-specific expression, they are also useful for the study of developmental gene expression. Promoters that drive cell-specific expression late in development (postnatally or in adult tissues only) are often used to define the functional role of a transgene product in adult tissue without the confounding expression of the transgene in embryonic tissues, which may produce effects arising from developmental alterations. Promoter truncation studies are also useful for identifying genetic enhancers and silencers, as well as hormone regulatory elements. 7.2.4.2 Dominant Mutants Expression of disease-associated mutations identified from human pedigree analyses can reveal the molecular basis for disease pathology. Mutations may be expressed either as a knockin as described earlier, with appropriate spatial and temporal regulation. However, the approach is technically labor intensive. Alternatively, dominant mutants can be studied by expressing the mutation from a tissue-specific promoter by injection into embryos derived from wild-type mice. This approach yields results more rapidly than by homologous recombination and does not require that isogenic DNA be isolated. This approach often produces a disease phenotype that appropriately models features seen in the human disease. Despite differences in gene expression level, the same phenotypic trend is observed with differences in severity of phenotype. Because the mutant transgene is usually expressed on a wild-type genetic background, the endogenous protein product is also present. For transgenic mutations that produce partially functional gene products, the normal endogenous protein may compensate for the defect. Genetic mutations in numerous genes, including rhodopsin (37,46–49), rds/peripherin (50), apoptotic genes (51,52) and others (53,54) have been studied using this methodology. Alternatively, dominant mutations may be introduced by pronuclear microinjection into embryos derived from null mutant mice, if such mice exist for the gene of interest. This can be somewhat riskier than creating the mice on a wild-type background and backcrossing onto the null mutant genetic background, as different strains of mice vary in their suitability to generate embryos for microinjection. As a rule of thumb, null mutant mice on a hybrid genetic background typically are more resilient to microinjection than are mice on a pure-bred genetic background. 7.2.4.3 Wild-Type Gene Expression Expression of the WT gene in a null mutant mouse line or a spontaneously occurring mutant mouse strain can also be used to verify the identity of a candidate gene, resulting in rescue of the pathological phenotype. For instance, a mutation in the phosphodiesterase β-subunit gene was confirmed as the cause of degeneration in the rd mouse by expressing the wild-type gene from the rod opsin promoter in the rd mouse line (19). Degeneration was also rescued in the rds (50) (retinal degeneration slow) mouse line by expression of the rds/peripherin gene. In the case where isoforms arise from alternative splicing, it can be useful to study isoform specific functions by the expression of specific splice isoforms in tissues normally expressing the isoform. For instance, the difference in the long and

7315.indb 154

11/15/07 1:40:31 PM


155

short forms of retinal arrestin were examined by microinjection of one of the forms onto a null arrestin genetic background (55). Alternatively, the effect of promiscuous stimulation of signaling or regulatory pathways may be studied by introducing a transgene that activates a biochemical pathway. For instance, persistent activation of the rod G-protein transducin causes degeneration (56). 7.2.4.4 Loss-of-Function Mutations Introduction of a null mutation is the most straightforward type of loss-of-function mutation, already described earlier. However, the production of a null mutation by homologous recombination in embryonic stem cells is labor intensive. If the gene to be investigated belongs to a characterized gene family or contains functional elements that have been defined in other genes, it is often possible to design dominant negative mutations that will impair gene function and interfere with the function of the normal endogenous protein. For instance, mutations can be made in domains that are known to be required for enzymatic activity, but retain a binding domain that competes with the normal endogenous protein. When little is known about functional domains of the gene, a comparison of the gene across different species can often identify highly conserved amino acids or domains that are likely to play an important role in gene function. Making such mutations can be useful for defining a specific functional role. An excellent example of such a study involved mutation of the phosphorylation sites of the rhodopsin gene (57,58). Such phosphorylation sites are highly conserved in the large family of G-protein-coupled receptors and has provided much information about the mechanism by which rhodopsin activity is terminated by phosphorylation. Again, pronuclear injection of mutants will produce founders with a range of expression levels. It is useful to correlate the level of transgene expression with the severity of phenotype.

7.2.5 Conditional or Regulated Transgene Expression Quite often, a gene product is expressed in more than one tissue type. In such cases, a null mutation or knockin mutation results in lethality, preventing study of gene function. Transgenic methods have been developed that allow “conditional” expression of the transgene in selected tissues. The simplest conditional knockouts are produced using a Cre/LoxP (59,60) (figure 7.5) or Flp/Frt system (61). For the null mutant gene to be produced, LoxP sites are introduced by homologous recombination into introns flanking a critical exon, creating what is referred to as a “floxed” gene. Embryonic stem cell clones containing the floxed gene are microinjected into blastocysts to produce transgenic mice. These mice are phenotypically normal until excision of DNA between the LoxP sites is induced by the expression of Cre recombinase. The Cre recombinase is expressed in a tissue-specific manner by producing a second mouse that expresses Cre recombinase from a tissue-specific promoter. This mouse is typically produced by pronuclear microinjection. Cross-breeding transgenic mice carrying the floxed gene with transgenic mice with tissue-specific expression of Cre recombinase produces transgenic mice in which a null mutation is present only in Cre-expressing tissues. This approach circumvents the confounding contribution of multisystemic

7315.indb 155

11/15/07 1:40:32 PM

156


Wild Type Gene 5' 1 Targeting Construct 5' 1 Targeted Gene before Cre Recombinase 5' 1 Targeted Gene after Cre Recombinase 5' 1

2

3

LoxP Positive Drug Marker

5

4

5

4

5

2

3

4

5

LoxP 2

3 LoxP


Floxed Gene in Transgenic

3'

LoxP

LoxP Positive Drug Marker

4

Negative Marker

3'

3'

3'

Tissue-specific Promoter/ Cre Recombinase Transgenic Mouse

Figure 7.5 Gene-targeting strategy for a conditional knockout using a Cre-LoxP system. LoxP sites and a positive drug-selectable marker such as neomycin, hygromycin, or puromycin are introduced by homologous recombination. A negative selectable marker such as thymidine kinase or diphtheria toxin may be attached to the 5′ or 3′ end of the targeting construct. A conditional knockout is achieved by crossing the mouse carrying the floxed gene with a second transgenic mouse expressing Cre recombinase driven from a tissue-specific promoter. This yields offspring that have the excised gene only in the targeted tissue type.

effects on other tissues. Transgenic mice targeting Cre recombinase to either rod (62–64) or cone (65,66) photoreceptor or other retinal cell types (67–70) have been developed. This methodology has been applied to the study of kinesin-2 function in photoreceptor cells (64). Transgene expression may also be regulated by introducing response elements regulated by tetracycline (71,72), rapamycin (73), progesterone (74) antagonist, or the insect hormone ecdysone (75). For example, these regulatable response elements can be incorporated into the Cre recombinase transgene construct and allow regulation of the timing and level of expression in a dose-dependent manner (76–78).

7.3 Conclusions We have provided a brief overview of the most commonly used strategies for the production of transgenic mouse models. The itemization is far from complete and combinations of the different features of each strategy are constantly evolving. In selecting a strategy to use, it is most important to carefully assess how stringently the animal model that you produce will answer the experimental questions proposed. The type of strategy taken for genetic mutations associated with dominantly inherited human diseases will be quite different from a strategy to study a highly

7315.indb 156

11/15/07 1:40:32 PM


157

conserved element (binding site, modification site, etc.) that has not been linked to human disease, but nevertheless is likely to play an important role in gene function. For the former, regulation of the transgene from the endogenous gene is less important than it is for the latter, where the physiological effects may be subtler. The availability of time, and correspondingly money, are also considerations. Although it may be desirable to knockin a mutation to study its function, the production of a targeting vector and homologous recombinant ES cell clones to the first generation of phenotypable offspring can take as long as 9 to 12 months, depending upon the complexity of the targeting design. For pronuclear injections, construction of the transgene to the initial analysis of mouse phenotype may take from 5 to 7 months. Thus, producing a knockout or knockin mouse takes nearly twice as long as producing a mouse by standard pronuclear microinjection. Regardless, with careful planning, the benefits of having a transgenic mouse model are a powerful tool that can provide unique information about in vivo gene function.

Acknowledgments JL acknowledges financial support from the National Eye Institute, Research to Prevent Blindness, Massachusetts Lions Eye Research Fund, and the Foundation Fighting Blindness. Thanks to Katherine Malanson for helpful comments on the manuscript.

References

7315.indb 157

1. Nagy, A., Gertsenstein, M., Vintersten, K., and Behringer, R. Manipulating the mouse embryo: A laboratory manual (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, 2003). 2. Ramirez-Solis, R., Davis, A.C., and Bradley, A. Gene targeting in embryonic stem cells. Methods Enzymol 225, 855–891 (1993). 3. Jackson, I.J. and Abbott, C.M. Mouse genetics and transgenics: a practical approach (Oxford University Press, Oxford, New York, 2000). 4. Hofker, M.H. and van Deursen, J. Transgenic mouse methods protocols (Humana Press, Totowa, NJ, 2003). 5. Brinster, R.L., Allen, J.M., Behringer, R.R., Gelinas, R.E., and Palmiter, R.D. Introns increase transcriptional efficiency in transgenic mice. Proc Natl Acad Sci U.S.A. 85, 836–840 (1988). 6. Qin, Y., Kong, L.K., Poirier, C., Truong, C., Overbeek, P.A., and Bishop, C.E. Long-range activation of Sox9 in Odd Sex (Ods) mice. Hum Mol Genet 13, 1213–1218 (2004). 7. Carter, D., Chakalova, L., Osborne, C.S., Dai, Y.F., and Fraser, P. Long-range chromatin regulatory interactions in vivo. Nat Genet 32, 623–626 (2002). 8. Calvert, P.D., Krasnoperova, N.V., Lyubarsky, A.L., Isayama, T., Nicolo, M., Kosaras, B., Wong, G., Gannon, K.S., Margolskee, R.F., Sidman, R.L., Pugh, E.N., Jr., Makino, C.L., and Lem, J. Phototransduction in transgenic mice after targeted deletion of the rod transducin alpha-subunit. Proc Natl Acad Sci U.S.A. 97, 13913–13918 (2000). 9. Xu, J., Dodd, R.L., Makino, C.L., Simon, M.I., Baylor, D.A., and Chen, J. Prolonged photoresponses in transgenic mouse rods lacking arrestin. Nature 389, 505–509 (1997). 10. Chen, C.K., Burns, M.E., He, W., Wensel, T.G., Baylor, D.A., and Simon, M.I. Slowed recovery of rod photoresponse in mice lacking the GTPase accelerating protein RGS9-1. Nature 403, 557–560 (2000).

11/15/07 1:40:33 PM

158


11. Chen, C.K., Burns, M.E., Spencer, M., Niemi, G.A., Chen, J., Hurley, J.B., Baylor, D.A., and Simon, M.I. Abnormal photoresponses and light-induced apoptosis in rods lacking rhodopsin kinase. Proc Natl Acad Sci U.S.A. 96, 3718–3722 (1999). 12. Humphries, M.M., Rancourt, D., Farrar, G.J., Kenna, P., Hazel, M., Bush, R.A., Sieving, P.A., Sheils, D.M., McNally, N., Creighton, P., Erven, A., Boros, A., Gulya, K., Capecchi, M.R., and Humphries, P. Retinopathy induced in mice by targeted disruption of the rhodopsin gene. Nat Genet 15, 216–219 (1997). 13. Liou, G.I., Fei, Y., Peachey, N.S., Matragoon, S., Wei, S., Blaner, W.S., Wang, Y., Liu, C., Gottesman, M.E., and Ripps, H. Early onset photoreceptor abnormalities induced by targeted disruption of the interphotoreceptor retinoid-binding protein gene. J Neurosci 18, 4511–4520 (1998). 14. Lem, J., Krasnoperova, N.V., Calvert, P.D., Kosaras, B., Cameron, D.A., Nicolo, M., Makino, C.L., and Sidman, R.L. Morphological, physiological, and biochemical changes in rhodopsin knockout mice. Proc Natl Acad Sci U.S.A. 96, 736–741 (1999). 15. Mendez, A., Burns, M.E., Sokal, I., Dizhoor, A.M., Baehr, W., Palczewski, K., Baylor, D.A., and Chen, J. Role of guanylate cyclase-activating proteins (GCAPs) in setting the flash sensitivity of rod photoreceptors. Proc Natl Acad Sci U.S.A. 98, 9948–9953 (2001). 16. Ramamurthy, V., Niemi, G.A., Reh, T.A., and Hurley, J.B. Leber congenital amaurosis linked to AIPL1: a mouse model reveals destabilization of cGMP phosphodiesterase. Proc Natl Acad Sci U.S.A. 101, 13897–13902 (2004). 17. Maeda, A., Maeda, T., Imanishi, Y., Kuksa, V., Alekseev, A., Bronson, J.D., Zhang, H., Zhu, L., Sun, W., Saperstein, D.A., Rieke, F., Baehr, W., and Palczewski, K. Role of photoreceptor-specific retinol dehydrogenase in the retinoid cycle in vivo. J Biol Chem 280, 18822–18832 (2005). 18. Wang, Z., Wen, X.H., Ablonczy, Z., Crouch, R.K., Makino, C.L., and Lem, J. Enhanced shutoff of phototransduction in transgenic mice expressing palmitoylation-deficient rhodopsin. J Biol Chem 280, 24293–24300 (2005). 19. Lem, J., Flannery, J.G., Li, T., Applebury, M.L., Farber, D.B., and Simon, M.I. Retinal degeneration is rescued in transgenic rd mice by expression of the cGMP phosphodiesterase beta subunit. Proc Natl Acad Sci U.S.A. 89, 4422–4426 (1992). 20. Turksen, K. Embryonic stem cells: methods and protocols. In Methods in Molecular Biology, Vol. 185, 499. ed. Turksen, K. Humana Press, Totowa, 2002. 21. Poueymirou, W.T., Auerbach, W., Frendewey, D., Hickey, J.F., Escaravage, J.M., Esau, L., Dore, A.T., Stevens, S., Adams, N.C., Dominguez, M.G., Gale, N.W., Yancopoulos, G.D., DeChiara, T.M., and Valenzuela, D.M. F0 generation mice fully derived from gene-targeted embryonic stem cells allowing immediate phenotypic analyses. Nat Biotechnol 25, 91–99 (2007). 22. te Riele, H., Maandag, E.R., and Berns, A. Highly efficient gene targeting in embryonic stem cells through homologous recombination with isogenic DNA constructs. Proc Natl Acad Sci U.S.A. 89, 5128–5132 (1992). 23. Ohtsuka, M., Ishii, K., Kikuti, Y.Y., Warita, T., Suzuki, D., Sato, M., Kimura, M., and Inoko, H. Construction of mouse 129/Ola BAC library for targeting experiments using E14 embryonic stem cells. Genes Genet Syst 81, 143–146 (2006). 24. Adams, D.J., Quail, M.A., Cox, T., van der Weyden, L., Gorick, B.D., Su, Q., Chan, W.I., Davies, R., Bonfield, J.K., Law, F., Humphray, S., Plumb, B., Liu, P., Rogers, J., and Bradley, A. A genome-wide, end-sequenced 129Sv BAC library resource for targeting vector construction. Genomics 86, 753–758 (2005). 25. Hasty, P., Rivera-Perez, J., and Bradley, A. The length of homology required for gene targeting in embryonic stem cells. Mol Cell Biol 11, 5586–5591 (1991). 26. Hasty, P., Rivera-Perez, J., Chang, C., and Bradley, A. Target frequency and integration pattern for insertion and replacement vectors in embryonic stem cells. Mol Cell Biol 11, 4509–4517 (1991).

7315.indb 158

11/15/07 1:40:33 PM


159

27. Tsang, S.H., Gouras, P., Yamashita, C.K., Kjeldbye, H., Fisher, J., Farber, D.B., and Goff, S.P. Retinal degeneration in mice lacking the gamma subunit of the rod cGMP phosphodiesterase. Science 272, 1026–1029 (1996). 28. Liang, Y., Fotiadis, D., Maeda, T., Maeda, A., Modzelewska, A., Filipek, S., Saperstein, D.A., Engel, A., and Palczewski, K. Rhodopsin signaling and organization in heterozygote rhodopsin knockout mice. J Biol Chem 279, 48189–48196 (2004). 29. Snove, O., Jr. and Rossi, J.J. Expressing short hairpin RNAs in vivo. Nat Methods 3, 689–695 (2006). 30. Snove, O., Jr. and Rossi, J.J. Toxicity in mice expressing short hairpin RNAs gives new insight into RNAi. Genome Biol 7, 231 (2006). 31. Singer, O., Tiscornia, G., Ikawa, M., and Verma, I.M. Rapid generation of knockdown transgenic mice by silencing lentiviral vectors. Nat Protocols 1, 286–292 (2006). 32. Campochiaro, P.A. Potential applications for RNAi to probe pathogenesis and develop new treatments for ocular disorders. Gene Ther 13, 559–562 (2006). 33. Cashman, S.M., Binkley, E.A., and Kumar-Singh, R. Towards mutation-independent silencing of genes involved in retinal degeneration by RNA interference. Gene Ther 12, 1223–1228 (2005). 34. Tessitore, A., Parisi, F., Denti, M.A., Allocca, M., Di Vicino, U., Domenici, L., Bozzoni, I., and Auricchio, A. Preferential silencing of a common dominant rhodopsin mutation does not inhibit retinal degeneration in a transgenic model. Mol Ther 14, 692–699 (2006). 35. Kiang, A.S., Palfi, A., Ader, M., Kenna, P.F., Millington-Ward, S., Clark, G., Kennan, A., O’Reilly, M., Tam, L.C., Aherne, A., McNally, N., Humphries, P., and Farrar, G.J. Toward a gene therapy for dominant disease: validation of an RNA interference-based mutation-independent approach. Mol Ther 12, 555–561 (2005). 36. Palfi, A., Ader, M., Kiang, A.S., Millington-Ward, S., Clark, G., O’Reilly, M., McMahon, H.P., Kenna, P.F., Humphries, P., and Farrar, G.J. RNAi-based suppression and replacement of rds-peripherin in retinal organotypic culture. Hum Mutat 27, 260–268 (2006). 37. Olsson, J.E., Gordon, J.W., Pawlyk, B.S., Roof, D., Hayes, A., Molday, R.S., Mukai, S., Cowley, G.S., Berson, E.L., and Dryja, T.P. Transgenic mice with a rhodopsin mutation (Pro23His): a mouse model of autosomal dominant retinitis pigmentosa. Neuron 9, 815–830 (1992). 38. Kassai, H., Aiba, A., Nakao, K., Nakamura, K., Katsuki, M., Xiong, W.H., Yau, K.W., Imai, H., Shichida, Y., Satomi, Y., Takao, T., Okano, T., and Fukada, Y. Farnesylation of retinal transducin underlies its translocation during light adaptation. Neuron 47, 529–539 (2005). 39. Vasireddy, V., Uchida, Y., Salem, N., Jr., Kim, S.Y., Mandal, M.N., Reddy, G.B., Bodepudi, R., Alderson, N.L., Brown, J.C., Hama, H., Dlugosz, A., Elias, P.M., Holleran, W.M., and Ayyagari, R. Loss of functional ELOVL4 depletes very long-chain fatty acids (>=C28) and the unique {omega}-O-acylceramides in skin leading to neonatal death. Hum Mol Genet 16, 471–482 (2007). 40. Lentz, J., Pan, F., Ng, S.S., Deininger, P., and Keats, B. Ush1c216A knock-in mouse survives Katrina. Mutat Res 616, 139–144 (2007). 41. Vasireddy, V., Jablonski, M.M., Mandal, M.N., Raz-Prag, D., Wang, X.F., Nizol, L., Iannaccone, A., Musch, D.C., Bush, R.A., Salem, N., Jr., Sieving, P.A., and Ayyagari, R. Elovl4 5-bp-deletion knock-in mice develop progressive photoreceptor degeneration. Invest Ophthalmol Vis Sci 47, 4558–4568 (2006). 42. Gross, A.K., Decker, G., Chan, F., Sandoval, I.M., Wilson, J.H., and Wensel, T.G. Defective development of photoreceptor membranes in a mouse model of recessive retinal degeneration. Vision Res 46, 4510–4518 (2006). 43. Chan, F., Bradley, A., Wensel, T.G., and Wilson, J.H. Knock-in human rhodopsin-GFP fusions as mouse models for human disease and targets for gene therapy. Proc Natl Acad Sci U.S.A. 101, 9109–9114 (2004).

7315.indb 159

11/15/07 1:40:33 PM

160


44. Weber, B.H., Lin, B., White, K., Kohler, K., Soboleva, G., Herterich, S., Seeliger, M.W., Jaissle, G.B., Grimm, C., Reme, C., Wenzel, A., Asan, E., and Schrewe, H. A mouse model for Sorsby fundus dystrophy. Invest Ophthalmol Vis Sci 43, 2732–2740 (2002). 45. Errijgers, V., Van Dam, D., Gantois, I., Van Ginneken, C.J., Grossman, A.W., D’Hooge, R., De Deyn, P.P., and Kooy, R.F. FVB.129P2-Pde6b(+) Tyr(c-ch)/Ant, a sighted variant of the FVB/N mouse strain suitable for behavioral analysis. Genes Brain Behav (2006). 46. Naash, M.I., Hollyfield, J.G., al-Ubaidi, M.R., and Baehr, W. Simulation of human autosomal dominant retinitis pigmentosa in transgenic mice expressing a mutated murine opsin gene. Proc Natl Acad Sci U.S.A. 90, 5499–5503 (1993). 47. Li, T., Franson, W.K., Gordon, J.W., Berson, E.L., and Dryja, T.P. Constitutive activation of phototransduction by K296E opsin is not a cause of photoreceptor degeneration. Proc Natl Acad Sci U.S.A. 92, 3551–3555 (1995). 48. Li, T., Snyder, W.K., Olsson, J.E., and Dryja, T.P. Transgenic mice carrying the dominant rhodopsin mutation P347S: evidence for defective vectorial transport of rhodopsin to the outer segments. Proc Natl Acad Sci U.S.A. 93, 14176–14181 (1996). 49. Li, T., Sandberg, M.A., Pawlyk, B.S., Rosner, B., Hayes, K.C., Dryja, T.P., and Berson, E.L. Effect of vitamin A supplementation on rhodopsin mutants threonine-17 --> methionine and proline-347 --> serine in transgenic mice and in cell cultures. Proc Natl Acad Sci U.S.A. 95, 11933–11938 (1998). 50. Travis, G.H., Groshan, K.R., Lloyd, M., and Bok, D. Complete rescue of photoreceptor dysplasia and degeneration in transgenic retinal degeneration slow (rds) mice. Neuron 9, 113–119 (1992). 51. Joseph, R.M. and Li, T. Overexpression of Bcl-2 or Bcl-XL transgenes and photoreceptor degeneration. Invest Ophthalmol Vis Sci 37, 2434–2446 (1996). 52. Nir, I., Kedzierski, W., Chen, J., and Travis, G.H. Expression of Bcl-2 protects against photoreceptor degeneration in retinal degeneration slow (rds) mice. J Neurosci 20, 2150–2154 (2000). 53. Salchow, D.J., Gouras, P., Doi, K., Goff, S.P., Schwinger, E., and Tsang, S.H. A point mutation (W70A) in the rod PDE-gamma gene desensitizing and delaying murine rod photoreceptors. Invest Ophthalmol Vis Sci 40, 3262–3267 (1999). 54. Olshevskaya, E.V., Calvert, P.D., Woodruff, M.L., Peshenko, I.V., Savchenko, A.B., Makino, C.L., Ho, Y.S., Fain, G.L., and Dizhoor, A.M. The Y99C mutation in guanylyl cyclase-activating protein 1 increases intracellular Ca2+ and causes photoreceptor degeneration in transgenic mice. J Neurosci 24, 6078–6085 (2004). 55. Burns, M.E., Mendez, A., Chen, C.K., Almuete, A., Quillinan, N., Simon, M.I., Baylor, D.A., and Chen, J. Deactivation of phosphorylated and nonphosphorylated rhodopsin by arrestin splice variants. J Neurosci 26, 1036–1044 (2006). 56. Raport, C.J., Lem, J., Makino, C., Chen, C.K., Fitch, C.L., Hobson, A., Baylor, D., Simon, M.I., and Hurley, J.B. Downregulation of cGMP phosphodiesterase induced by expression of GTPase-deficient cone transducin in mouse rod photoreceptors. Invest Ophthalmol Vis Sci 35, 2932–2947 (1994). 57. Mendez, A., Burns, M.E., Roca, A., Lem, J., Wu, L.W., Simon, M.I., Baylor, D.A., and Chen, J. Rapid and reproducible deactivation of rhodopsin requires multiple phosphorylation sites. Neuron 28, 153–164 (2000). 58. Doan, T., Mendez, A., Detwiler, P.B., Chen, J., and Rieke, F. Multiple phosphorylation sites confer reproducibility of the rod’s single-photon responses. Science 313, 530–533 (2006). 59. Yu, Y. and Bradley, A. Engineering chromosomal rearrangements in mice. Nat Rev Genet 2, 780–790 (2001). 60. Kos, C.H. Cre/loxP system for generating tissue-specific knockout mouse models. Nutr Rev 62, 243–246 (2004).

7315.indb 160

11/15/07 1:40:34 PM


161

61. Marszalek, J.R., Liu, X., Roberts, E.A., Chui, D., Marth, J.D., Williams, D.S., and Goldstein, L.S. Genetic evidence for selective transport of opsin and arrestin by kinesin-II in mammalian photoreceptors. Cell 102, 175–187 (2000). 62. Le, Y.Z., Zheng, L., Zheng, W., Ash, J.D., Agbaga, M.P., Zhu, M., and Anderson, R.E. Mouse opsin promoter-directed Cre recombinase expression in transgenic mice. Mol Vis 12, 389–398 (2006). 63. Li, S., Chen, D., Sauve, Y., McCandless, J., Chen, Y.J., and Chen, C.K. RhodopsiniCre transgenic mouse line for Cre-mediated rod-specific gene targeting. Genesis 41, 73–80 (2005). 64. Jimeno, D., Feiner, L., Lillo, C., Teofilo, K., Goldstein, L.S., Pierce, E.A., and Williams, D.S. Analysis of kinesin-2 function in photoreceptor cells using synchronous Cre-loxP knockout of Kif3a with RHO-Cre. Invest Ophthalmol Vis Sci 47, 5039–5046 (2006). 65. Le, Y.Z., Ash, J.D., Al-Ubaidi, M.R., Chen, Y., Ma, J.X., and Anderson, R.E. Targeted expression of Cre recombinase to cone photoreceptors in transgenic mice. Mol Vis 10, 1011–1018 (2004). 66. Akimoto, M., Filippova, E., Gage, P.J., Zhu, X., Craft, C.M., and Swaroop, A. Transgenic mice expressing Cre-recombinase specifically in M- or S-cone photoreceptors. Invest Ophthalmol Vis Sci 45, 42–47 (2004). 67. Matsuda, T. and Cepko, C.L. Controlled expression of transgenes introduced by in vivo electroporation. Proc Natl Acad Sci U.S.A. 104, 1027–1032 (2007). 68. Zhang, X.M., Chen, B.Y., Ng, A.H., Tanner, J.A., Tay, D., So, K.F., Rachel, R.A., Copeland, N.G., Jenkins, N.A., and Huang, J.D. Transgenic mice expressing Cre-recombinase specifically in retinal rod bipolar neurons. Invest Ophthalmol Vis Sci 46, 3515–3520 (2005). 69. Gelman, D.M., Noain, D., Avale, M.E., Otero, V., Low, M.J., and Rubinstein, M. Transgenic mice engineered to target Cre/loxP-mediated DNA recombination into catecholaminergic neurons. Genesis 36, 196–202 (2003). 70. Campsall, K.D., Mazerolle, C.J., De Repentingy, Y., Kothary, R., and Wallace, V.A. Characterization of transgene expression and Cre recombinase activity in a panel of Thy-1 promoter-Cre transgenic mice. Dev Dyn 224, 135–143 (2002). 71. Morgan, W.W., Richardson, A., Sharp, Z.D., and Walter, C.A. Application of exogenously regulatable promoter systems to transgenic models for the study of aging. J Gerontol A Biol Sci Med Sci 54, B30–40; discussion B41–42 (1999). 72. Albanese, C., Hulit, J., Sakamaki, T., and Pestell, R.G. Recent advances in inducible expression in transgenic mice. Semin Cell Dev Biol 13, 129–141 (2002). 73. Harvey, D.M. and Caskey, C.T. Inducible control of gene expression: prospects for gene therapy. Curr Opin Chem Biol 2, 512–518 (1998). 74. Tsai, S.Y., O’Malley, B.W., DeMayo, F.J., Wang, Y., and Chua, S.S. A novel RU486 inducible system for the activation and repression of genes. Adv Drug Deliv Rev 30, 23–31 (1998). 75. No, D., Yao, T.P., and Evans, R.M. Ecdysone-inducible gene expression in mammalian cells and transgenic mice. Proc Natl Acad Sci U.S.A. 93, 3346–3351 (1996). 76. Hasan, M.T., Friedrich, R.W., Euler, T., Larkum, M.E., Giese, G., Both, M., Duebel, J., Waters, J., Bujard, H., Griesbeck, O., Tsien, R.Y., Nagai, T., Miyawaki, A., and Denk, W. Functional fluorescent Ca2+ indicator proteins in transgenic mice under TET control. PLoS Biol 2, e163 (2004). 77. Okoye, G., Zimmer, J., Sung, J., Gehlbach, P., Deering, T., Nambu, H., Hackett, S., Melia, M., Esumi, N., Zack, D.J., and Campochiaro, P.A. Increased expression of brain-derived neurotrophic factor preserves retinal function and slows cell death from rhodopsin mutation or oxidative damage. J Neurosci 23, 4164–4172 (2003). 78. Chang, M.A., Horner, J.W., Conklin, B.R., DePinho, R.A., Bok, D., and Zack, D.J. Tetracycline-inducible system for photoreceptor-specific gene expression. Invest Ophthalmol Vis Sci 41, 4281–4287 (2000).

7315.indb 161

11/15/07 1:40:34 PM

7315.indb 162

11/15/07 1:40:34 PM

2 Vertebrate Nonvisual Phototransduction

7315.indb 163

11/15/07 1:40:48 PM

7315.indb 164

11/15/07 1:40:48 PM

8

Melanopsin Signaling and Nonvisual Ocular Photoreception Sowmya V. Yelamanchili, Victoria Piamonte, Surendra Kumar Nayak, Nobushige Tanaka, Quansheng Zhu, Kacee Jones, Hiep Le, and Satchidananda Panda

Contents 8.1 8.2

Introduction.................................................................................................. 165 Animal Behavioral Studies and Expression Patterns.................................. 167 8.2.1 Pupillary Light Response (PLR)....................................................... 168 8.2.1 Circadian Photoentrainment Assay.................................................. 170 8.2.2 Photic Suppression of Pineal Melatonin Synthesis and Release...... 173 8.3 Identification and Characterization of Melanopsin-Expressing Cells......... 173 8.4 Heterologous Expression and Functional Assays......................................... 177 8.4.1 Mammalian Expression System....................................................... 178 8.4.2 Melanopsin FLIPR Assay................................................................. 179 8.4.3 Functional Expression of Melanopsin in Xenopus Oocyte Expression System............................................................................ 182 8.5 Purification of Melanopsin for Spectral and Biochemical Studies.............. 187 References............................................................................................................... 189

8.1 Introduction Organisms, ranging from cyanobacteria to mammals, use an intrinsic circadian clock and photosensors to regulate behavior and physiology in resonance with geophysical time. In mammals, an endogenous master clock functions in the suprachiasmatic nuclei (SCN) of the hypothalamus, and is entrained to the day–night cycle by direct light input from the retina. Appropriate photoentrainment is essential for general health, as shift work and similar chronic circadian desynchronization have been 165

7315.indb 165

11/19/07 8:39:43 PM

166


shown to be major risk factors in several sleep disorders, metabolic syndromes, and in cancer (reviewed in (1)). Understanding the molecular processes underlying photoentrainment will therefore help identify new strategies and targets for therapeutic intervention in these disorders. In addition to circadian photoentrainment, light received through mammalian eyes also mediates several other photoadaptive behaviors and physiologies. These include dynamic pupillary light responses (PLRs), and light modulation of the neuroendocrine system such as light-induced suppression of pineal melatonin synthesis and release. Finally, ocular photoresponse also accounts for temporal niche-dependent modulation of activity-rest state, i.e., general suppression of activity in nocturnal species, and improvement in alertness in diurnal mammals. Persistence of these photic responses in many blind human patients and in blind animal models deficient in rod/cone signaling components or with outer retinal degeneration has led to a collective description of these responses as nonvisual, non-image-forming, or adaptive photic responses [reviewed in Van Gelder (2)]. Complete loss of these responses in experimental bilateral enucleation animal studies established that a novel nonrod, noncone inner retina photopigment may play a significant role in adaptive light responses. The spectral nature of the novel photopigment was predicted from the action spectra of circadian photoentrainment and PLR assays in rodents (3,4). A novel combination of retrograde labeling to mark retinal ganglion cells (RGCs) that directly project to the circadian brain center, the SCN, and electrophysiological characterization of light responsiveness of these neurons led to the discovery of intrinsically photosensitive retinal ganglion cells (ipRGCs) (5,6). The ipRGCs demonstrated some unique properties distinct from those of rod/cone photoreceptors: light-evoked membrane depolarization, high threshold sensitivity, long response latency, slow deactivation, and resistance to bleaching (5). Homology cloning of melanopsin (7), colocalization of the protein product with ipRGCs (6), and phenotypic analysis of melanopsin-deficient mice has established that melanopsin functions as a true photopigment in the ipRGCs and significantly contributes to adaptive photoresponses (6,8–11). The discovery of melanopsin’s role in adaptive photoresponses has opened an entry point to understand the molecular bases of nonvisual photoresponses. Genetic analysis has established that both rod/cone and melanopsin photopigments participate in these processes, albeit with varying magnitudes specific for the individual response (9,12). Currently, the relative contribution of rod or cone photoreceptors to various non-image-forming responses is unclear. The neuronal circuitry and the underlying molecular signaling events conveying light information from rod/cone and ipRGCs to the brain centers regulating different adaptive responses have yet to be clearly mapped out. In parallel, molecular and biochemical characterization of melanopsin photopigment will unravel which properties of the ipRGCs are encoded in the photopigment and how photoexcited melanopsin transmits light information along the downstream signaling cascade. Several methods and techniques at whole-animal, cellular, and biochemical levels combined with genetic and pharmacological perturbations will prove invaluable in addressing these questions. We will describe some techniques widely

7315.indb 166

11/19/07 8:39:44 PM

Melanopsin Signaling and Nonvisual Ocular Photoreception

167

used by several labs and a few recently developed in our lab. In each section we will also highlight how use of the specific approach or technique has made a seminal contribution to our understanding of adaptive photoresponses in mammals. The first half of the chapter will describe methods using whole-animal or animal tissue, and in the second part we will focus on heterologously expressed melanopsin.

8.2 Animal behavioral studies and expression patterns Photoadaptive behavior assays in small rodents have been extremely useful in defining the nature of the photopigments driving these responses, and genetically dissecting their relative contributions. The widely used assays include: (a) PLR, (b) light-induced suppression of activity in nocturnal rodents (or negative masking), (c) light-induced suppression of pineal melatonin synthesis and release, and (d) circadian photoentrainment. These four behaviors are regulated by different brain centers and differ in underlying neuronal signaling network, threshold sensitivity, dynamic range of sensitivity, and response under prolonged illumination. With careful planning, all four assays can be performed on the same group of animals. Pupillary responsiveness to light is a relatively sensitive and quick photoadaptive response in which the pupil constricts within milliseconds of sudden increase in irradiance and the constriction is maintained under prolonged illumination. The assay has an excellent dynamic range of sensitivity over several orders of irradiance levels, and the results are relatively independent of the time of the day the assay is performed. Similar to PLR, negative masking is an assay with significant dynamic range of sensitivity. Within a few seconds of bright illumination during the subjective night, mice usually reduce their overall activity and under prolonged illumination over hours the activity partially recovers. Light-induced suppression of pineal melatonin synthesis and release is yet another assay conducted during the subjective night, when the melatonin synthesis usually rises. Within several minutes of illumination, the pineal melatonin levels, and subsequently, circulating melatonin levels, drop. Finally, the circadian photoentrainment assay in rodents is a resourceintensive procedure that is often used to measure several aspects of the effect of light on circadian rhythms, including general entrainment to an imposed light–dark cycle, the phase-shifting effect of light on circadian behavior under constant darkness, and circadian behavior under constant light. The light sensitivity of these four behavioral responses appropriately reflects the relative contributions of the rod/cone pathway and melanopsin. A rapid response such as PLR at all light intensities is dominantly regulated by the rod/cone pathway, so that the loss of outer retina leads to a log unit reduction in photosensitivity of PLR, whereas the loss of melanopsin causes a modest change in PLR only at high light intensity. The melanopsin pathway, however, plays a dominant role in negative masking and circadian photoentrainment that exhibit high threshold sensitivity, where the loss of melanopsin causes a significant change in photic sensitivity of these responses, whereas complete outer retina degeneration does not cause any detectable change in sensitivity. Although outer retina degeneration causes no significant

7315.indb 167

11/19/07 8:39:44 PM

168


reduction in the sensitivity of photic suppression of pineal melatonin synthesis, the effect of the loss of melanopsin has not been extensively studied. Such differential sensitivities will allow fine molecular dissection of adaptive photoresponses using whole-animal behavior assays.

8.2.1 Pupillary Light Response (PLR) PLR is a simple yet powerful assay that has been widely used in characterizing photopigments and signaling components driving nonvisual photoresponses in animals. Persistence of PLR in mice with retina degeneration (rd) was the first observation suggesting the existence of nonvisual photoreceptors in mammals (13). Although the PLR in some birds (14) can be driven by iris-resident photopigment ex vivo, PLR in small rodents is usually performed in live anesthetized animals. The mouse is handheld by the scruff along an illuminated light path and a video image of the pupil is acquired for up to 3 min. For detailed descriptions of measuring mouse PLR, please refer to Van Gelder (15). Protocol 1: Measurement of PLR Instrument setup

7315.indb 168

1. Camera: A consumer-grade video camera with infrared recording ability is commonly used. The camera should be fitted with a macro and close-up lens. If the video image is to be processed by an image processing software such as Eye Tracker (Arrington Research, Arizona), outfitting the camera with an additional infrared filter (89C or 89B filter) usually improves the contrast. 2. Infrared light source: The built-in infrared LEDs in the video camera usually provide enough illumination. However, in some camera models, mounting additional filters may partially block the infrared LED. An extra infrared LED source, easily available from several Internet vendors, usually improves image quality. 3. Visible light source: A halogen or xenon light source fitted with a fiberoptic or liquid light guide. Either the light source or the light guide should be fitted with a filter holder to hold narrow band-pass interference and/or neutral density filters to regulate spectral quality. A calibrated photometer is used to measure irradiance level from various filter combinations. The light source is preferably fitted with a foot-pedal-activated shutter. 4. Ancillary equipment: We typically use several flexible arm magnetic bases, mounting screws to position and secure the video camera, infrared LED, light guide, and the irradiance detector on a piece of steel plate. The display monitor of the video camera is covered with a red acetate filter (Lee filter #029 or #106). The researcher uses a headlamp covered with the same red acetate filter to handle mice during the assay. All visible light leaks from the light source and other instruments in the room are blocked with opaque material following the institute’s safety guideline. 5. Ergonomics: All instruments and the camera lenses should be appropriately positioned and adjusted to obtain the best-quality infrared video of

11/19/07 8:39:44 PM


169

the mouse eye. We use some additional magnetic bases and flexible arms for proper positioning of the mouse without causing too much stress on the animal or the researcher. Pupil measurement

6. Familiarize the mice to human handling for up to 1 week prior to the experiment to reduce stress. 7. Dark-adapt the mice for at least 2 h before the commencement of the experiment so as to completely dilate the pupil. 8. Hold the mouse by its scruff and position its eye to a point where a sharp infrared video image is captured by the camera. Capture a dark baseline video for up to 1 min. 9. Illuminate the mouse eye and collect the video image for another 1 min. In wild-type mice under bright illumination, pupil constriction commences within a few milliseconds, and maximum constriction is achieved within 10 s. 10. After termination of the light pulse, record an additional 30 s to 1 min infrared video to measure pupil relaxation. 11. Measure repeatedly pupil constrictions on the same animal after ~2 h of dark adaptation between measurements. Change the interference filter and/ or neutral density filters after one batch of animals is examined at a given light quality. Examine the mice for their PLR to a different light quality after the dark adaptation. 12. Measure the pupil diameter of the enlarged video images on the display screen. Calculate the ratio of cornea diameter to pupil diameter. Alternatively, use automated image analysis software, such as the pupil measurement feature of Eye Tracker (Arrington Research) to measure pupil diameter at a defined interval. In our experience the image processing software is typically optimized for larger pupils and adapting them for the mouse eye needs a significant amount of optimization of the light, video quality, contrast, and image size.

Several factors including stress level can significantly interfere with PLR, thereby slowing the pupil constriction rate and preventing complete pupil constriction in wild-type mice even at high irradiance level. Therefore, noise level and human activity in the room should be kept to a minimum. Occasionally, mice show anterior chamber dysgenesis leading to attachment of the iris to the cornea, which causes incomplete constriction. In rd;opn4−/− and comparable genotypes with no rod/cone and melanopsin function, the pupil does not constrict even in response to bright light (9,12). To ensure the lack of pupil constriction is due to a deficiency in the light signaling pathway and not a result of abnormal development of the neuromusculature responsible for pupil constriction, at the end of PLR a drop of dilute cholinergic agonist such as 1% pilocarpine may be applied to each eye. The drug usually completely constricts the pupil within a few minutes of application.

7315.indb 169

11/19/07 8:39:44 PM

170


8.2.1 Circadian Photoentrainment Assay Circadian activity measurement in rodents has been extensively used to characterize photoreceptor properties (4,16) and for genetic dissection of light input pathway (9,12) that synchronizes the circadian clock to the ambient lighting. Action spectra of circadian photoentrainment in hamsters predicted that a vitamin-A-based photopigment with peak response of around 480–500 nm (4) may be involved. However, intact photoentrainment in rd mice (16) and in various other outer retina degeneration mouse models (17,18) established that rod/cone photoreceptors are dispensable. Normal entrainment of melanopsin-deficient mice to an imposed light–dark cycle and the attenuated phase shift of the oscillator in response to discrete light pulse demonstrated the complex interplay between the rod/cone and melanopsin pathways in photoentrainment, where the loss of melanopsin can be partially compensated by rod/cone photoreceptors whereas the loss of rod/cone function can be fully compensated by melanopsin. Complete loss of entrainment in mice deficient in rod/cone and melanopsin function proved the necessity and sufficiency of these photopigments in circadian photoentrainment (9,12). As genetic perturbation models targeting specific molecules or cell types become available, the photoentrainment assay will continue to be a powerful tool for examining this process. The assay also serves to monitor photic suppression of activity in nocturnal rodents. Negative masking behavior in mice deficient in rod/cone and/or melanopsin function closely parallels circadian behavior in these mice. The loss of rods/ cones have little to no effect on negative masking in response to bright light, whereas opn4−/− mice become more active than wild-type littermates under prolonged illumination (19). Mice with neither functional rod/cone nor melanopsin fail to show any negative masking (9,12). Although small-animal activity can be monitored by various means such as infrared beam breaks, temporal monitoring of drinking or eating, and body temperature telemetry, wheel running activity measurement often produces better quality data and is widely used in various circadian labs. For a thorough description of the room setup, apparatus, data collection and analysis software, we strongly encourage readers to refer to a recent article (20). Mice are individually housed in specially designed cages equipped with running wheels. Up to 24 wheel cages are placed inside a light-tight box with independent light control, ventilation, and sensors for humidity, temperature, and light. A constant level of humidity and temperature is maintained inside the room. The ventilation fans provide constant white noise that masks all other temporal auditory cues, such as routine human activities in the room. Individual wheel cages are connected via switches to the ClocklabTM acquisition software (Actimetrics, Chicago, Illinois) that continuously records wheel rotation over days and weeks. Wheel rotations are recorded in 1–10 min bins. Protocol 2: Circadian Wheel Running Assay

7315.indb 170

1. House mice individually in wheel running cages under 12 h light and 12 h dark (LD) during the first week of the experiment to entrain their activity

11/19/07 8:39:45 PM

171


8

12

16

Local Time 20

0

4

8 Light:Dark (LD)

rhythm to the imposed lighting regime. Mice of C57Bl/6 background usually consolidate their wheel running activity to the night, so that the onset of wheel running activity coincides with the dark onset. They run on the wheel almost continuously for the first 6–8 h of the night, after which the wheel running is intermittent for the rest of the night. Very little wheel running activity is observed during the day (figure 8.1). 2. For measurement of negative masking, switch on lights in the wheel running chamber for 1–3 h, starting from 2 h after the beginning of the dark phase to assess the effect of light on activity (if any) by comparing the activity in 1–10 min bins during the light pulse with activity during the comparable time for the previous 2–3 nights. 3. Because the light pulse used to assess photic suppression of activity can also reset the circadian oscillator, to ensure proper entrainment, mice are again returned to light–dark cycles for one additional week.

10 15

25 Light Pulse 30

Phase Shift

Constant Dark (DD)

Days

20

35

Figure 8.1 Light regulation of circadian wheel running activity of mouse. Wheel running activity record of a mouse showing entrainment to a light–dark cycle and phase shift in response to a brief pulse of light. Vertical bars along each horizontal line show wheel running activity over 24 h. Local time and period of light and dark during entrainment are represented by white and dark boxes on the top. The mouse was placed in a light–dark cycle for almost 2 weeks during which the circadian clock is entrained to the imposed LD cycle and the time of activity onset maintains a constant phase relationship with the time of dark onset. After 2 weeks of entrainment, the mouse was released into constant darkness. Under constant darkness, the activity rhythm is under the control of the endogenous circadian oscillator, which in this mouse runs with a period length of 500 bases. Incubate at 37°C for 2 h when making transcripts 95% genome coverage (cited in Reference 14). In addition, a large number of deficiencies whose breakpoints have not been determined are also available. However, the coverage is still quite uneven. Some regions are densely covered with deficiencies, whereas for some others, coverage is relatively poor. Moreover, for most of the deficiencies available, the breakpoints have not been determined at the molecular level. Thus, even with the improved coverage and improved molecular characterization of deficiency breakpoints, our experience has been that, with some exceptions, for most mutations of interest it is difficult to map reliably with a resolution of less than ~50–100 kb. This is still too large a region to identify the target gene readily. The recent improvements in mapping to be discussed in the next section are intended to improve the resolution further.

9.2.2 Recent Improvements in Mapping Strategy Several different mapping strategies that would considerably improve the resolution have been proposed recently. They are P-element-mediated male recombination mapping (15,16), SNP (single-nucleotide polymorphism)-based mapping (17), and molecularly defined P-element insertion-based mapping (11). P-element-mediated male recombination mapping relies on the fact that male recombination, which is normally rare in Drosophila, can occur at a frequency of up to 1% in crosses involving fly strains carrying P elements. Because recombination occurs primarily at the site of P element insertion (15), from the examination of recombinants identified with the help of flanking markers, one can determine if any given P element chosen for mapping lies to the right or left of the mutation. Thus, using a series of P elements in the region of interest, any given target mutation can be mapped to an accuracy that depends only on the density of P elements in the region (16). As in the case of deficiencies, there have been ongoing efforts to increase the coverage of the genome by P element insertions. The Bloomington Drosophila Stock Center lists nearly 10,000 P insertion lines that are mapped at the nucleotide level (Flybase) (http://www.flybase. org). Perhaps, the largest collection of molecularly defined P element insertion lines currently in existence is that generated by a group headed by Jaeseob Kim of Korea Advanced Institute of Science and Technology, Daejon, Korea, and marketed by GenExel-Sein, Inc. This group generated nearly 100,000 P insertion lines, ~25,000

7315.indb 198

11/15/07 1:42:11 PM

Phototransduction in Drosophila

199

of which are said to be unique. If one includes those that are inserted immediately upstream of a gene on the sense strand and those immediately downstream of a gene on the antisense strand,1 as well as those within the gene, the genome coverage is estimated to be about 52–70% (Eunkyung Bae, GenExel, private communication). Genome coverage by P element insertions is not expected to reach 100% because a substantial fraction of genes in the Drosophila genome is thought to be refractory to P insertions. Nevertheless, P element insertions are sufficiently abundant to make this an effective tool. However, it has a number of drawbacks (see also Reference 11) and does not eliminate many of the time-consuming steps in mapping. First of all, the method is not applicable to mutations on the X chromosome because it requires that the chromosome carrying the mutation and the homologous chromosome carrying the P element be in trans to each other in male flies. This is not possible, because male flies have only one X chromosome. Second, the method requires that visible markers be recombined into the chromosomes carrying the mutations being mapped. Third, to obtain sufficient accuracy, several rounds of mapping using chromosomes with different P element insertions need to be performed. Fourth, at each round of mapping, this method maps the mutation either to the left or right of the P insertion and does not provide any further clues to the approximate map position to guide the choice of the next P insertion. Finally, the method does not necessarily identify the gene carrying the mutation. It ultimately maps the mutation between two adjacent P insertions in the chromosome. In a region densely covered with P insertions, there may only be a small number of genes within the mapped interval. In a sparsely covered region, however, there could be tens of, or even over a hundred, genes in the interval. The second method is one based on SNPs (17). It is basically meiotic recombination mapping that utilizes SNPs as molecular markers for recombination events. SNPs are single-nucleotide differences between homologous chromosomes that have been shown to exist in sufficient numbers in the Drosophila genome for this method to be practicable. Crosses are set up to allow meiotic recombinations to take place between the chromosome bearing the mutation to be mapped and a homologous chromosome that bears a visible marker. The visible marker is often a P insertion into which a visible mutation has been engineered. Recombinations occur in females heterozygous for the aforementioned two chromosomes, and recombinants are detected in the following generation with the help of the visible marker. This method of mapping requires that one first construct an SNP map in the region of the chromosome to be tested. SNPs used for the map must be polymorphic between the marker chromosome and the mutant chromosome such that their chromosomal origin can be identified either from restriction fragment length polymorphisms or sequence differences. The recombinant offspring that are identified with the help of the P insertion marker are then tested for both their mutant phenotype (by ERG recording in our case) and the genotypes of the SNP markers. The genotypes of SNPs in any particular recombinant establish the site of the crossover event in relation to the SNP map, and the phenotype (mutant or wild type) of the recombinant establishes the site of the mutation in relation to the site of the crossover (to the left or right of the crossover event). If a sufficient number of such recombinants is analyzed, one can map the mutation between two neighboring SNPs in the SNP map. The recombinant events

7315.indb 199

11/15/07 1:42:12 PM

200


most informative regarding the mapping would be the rarest ones that occur between the mutation and the SNP nearest to it. Among the major drawbacks of this method is the necessity for constructing an SNP map each time a mutation in a different region of the genome is to be mapped. If preexisting information on SNPs is used in the construction of the map, it will still have to be confirmed. Moreover, the greater the density of SNP markers and the greater the number of recombinants generated and tested, the greater will be the accuracy of mapping, and each recombinant will have to be tested for the mutant phenotype by ERG recording and for the genotype of SNP markers by either restriction digestion or sequencing. All these are potential time-consuming steps. In addition, the final outcome, in general, will not necessarily be the identification of a single target gene but a number of candidate genes. Still another recently proposed method of mapping is that based on the use of molecularly defined P element insertions (11). This method is essentially classical meiotic recombination mapping that uses, instead of visible markers, P element insertions as markers to determine recombination rates. Potentially, this can be a powerful method because a large number of P insertions, for which the sites of insertions have been determined at the molecular level, are already available, and more are being generated (see previous section). A major drawback of this method of mapping, however, is that because of the large number of recombinants that need to be scored, its use is restricted to mapping mutations that produce easily scorable phenotypes. Zhai et al. (11), for example, used this method to map lethal mutations. They carried out their mapping in two steps, rough mapping followed by fine mapping. Both utilized the same recombination mapping strategy, but in fine mapping they used closely spaced P insertions that span the interval defined by rough mapping. They recommend scoring >1000 recombinant progeny for each P insertion cross to attain an accuracy of 200 kb region centered around the location of inaF, the next highest log-fold change detected by any probe set was −0.765, and the gene corresponding to this probe set was located about 100 kb away from inaF (data not shown). We thus concluded that, with null mutants, the microarray strategy will pinpoint the corresponding genes immediately and unequivocally. Mapping need not even be highly refined.

9.6.2 Previously Unidentified Genes 9.6.2.1 inaE However, if the microarray strategy were to be generally useful, we need to show that it can be successfully applied to previously unidentified genes. Moreover, most mutants are not null mutants and their phenotypes are often relatively mild. We therefore need to show that this strategy can be used in the isolation of genes in which only mild mutants have been isolated. We chose the inaE gene for the first “live test” of the strategy. The mutant inaEN125 is one of the two classically generated mutant alleles of the inaE gene (41), which we had been attempting to clone for some time with limited success. From the mutant phenotype, the inaE-encoded protein was expected to be involved in phototransduction. We felt inaEN125 was a very good test of the strategy because its phenotype is rather mild and, for technical reasons, only one of the two allelic mutants could be used for microarray analysis. Thus, it was expected to tax the efficacy of the strategy. The inaE gene was mapped to the 12C5-6;12C8-D1 region of the X chromosome by deficiency mapping. Two critical deficiencies for this mapping were Df(1)AR10 (12B1-2;12C8-D1) and Df(1)benCO2 (12C5-6;12E), both of which uncovered the mutation; i.e., they did not complement the mutation. Again, we searched a much broader region than that determined by mapping: 12B4 to 12D4, a ~350 kb region with 41 genes (table 9.2). In this region, probe sets for the following three genes detected log-fold expression changes greater than 0.3 in absolute value in the inaE mutant compared to wild type: Yp3 at 12C1, CG33174 at 12C4-5, and CG32626 at 12C6-7. CG33174 was of particular interest because this was the only gene in this region for which the changes in expression detected by its probe sets were assigned a P value of 0 (table 9.2). Southern blots of deficiency heterozygotes, Df(1)AR10/+ and Df(1) benCO2/+, probed with CG32626 probes showed that CG32626 is outside the mapped limits of inaE. The Yolk protein 3 gene (Yp3) had been previously characterized both genetically and molecularly (e.g., 42–44). We obtained two Yp3 mutants from Dr. Mary Bownes of the University of Edinburgh and carried out complementation tests between them and inaE. A Yp3 mutant allele when placed in trans with inaE complemented the inaE phenotype, definitively ruling out Yp3 as the inaE gene. Thus, the preceding analysis left CG33174 as the only remaining candidate gene for inaE. To determine whether CG33174 indeed is inaE, we carried out the following three independent lines of validation experiments (41): (a) sequencing the CG33174 gene in the inaEN125 mutant, (b) complementation tests between inaE and existing P-insertion mutations in CG33174, and (c) in vivo rescue of inaE by germline transformation of inaE with a wild-type copy of CG33174. Sequencing showed

7315.indb 207

11/15/07 1:42:15 PM

208


Table 9.2 Genes Showing Expression Changes in inaE (12B4-D4) Probe Set ID

Log-Fold Change

1641025_at 1631419_at 1639574_a_at 1637006_at 1627706_a_at 1625176_at 1640287_s_at 1631073_at 1636030_s_at 1632081_s_at 1638749_at 1632900_at 1625285_at 1630614_s_at

–0.086208 –0.64749 0.17604 0.13555 0.657 0.56423 0.49505 0.06994 0.13365 0.40171 0.13435 0.12029 0.10004 –0.18765

P Value

Alignments

Gene Symbol

0.010916 7.84E-05 6.75E-05 0.0061398 0 0 0 0.0079756 0.0015452 1.56E-09 0.0012134 0.002592 0.0057516 0.0003122

X:13468801-13469838 X:13490203-13491835 X:13493428-13507399 X:13514407-13536302 X:13514407-13536302 X:13514407-13542160 X:13540023-13541832 X:13544100-13552983 X:13568261-13568744 X:13574612-13576134 X:13600040-13616171 X:13632111-13632282 X:13640807-13642515 X:13737316-13737562

CG11134 Yp3 rdgB CG33174 CG33174 CG33174 CG33174 1(1)G0053 CG32626 CG32626 CG32611 CG11072

Cytol Position 12B6 12C1 12C1-2 12C4-5 12C4-5 12C4-5 12C5 12C5-6 12C6-7 12C6-7 12C7 12C8 12D1 12D2

Note: Again, only those probe sets satisfying the significance level of 0.05 (9.5 statistical analysis) are included in the table. The search covered the region from 12B4 to 12D4. However, the table shows probe sets in a slightly narrower region because the other probe sets did not satisfy the preceding criterion. The same comments as in table 9.1 regarding alignments apply.

that the inaE mutant carries a mutation at the 5′ splice site of the 11th intron. As a consequence, this intron fails to be spliced out, and translation proceeds into the intron sequence until a stop codon is encountered within the intron. The CG33174 gene had been characterized only electronically. There have been no classically generated mutants in this gene. However, there were four P insertion alleles listed in Flybase. These were obtained, and complementation tests were carried out between each of these and inaEN125. inaEN125 is characterized by two distinct ERG phenotypes. One is that its light-evoked response starts decaying during illumination, and the other is that it enhances (makes the phenotype stronger) the ERG phenotype of TrpP365/+. If the mutation is relatively mild, the first of these phenotypes can be troublesome to detect in a red-eye background, as is the case of P-insertion lines. In fact, none of the four P-insertion lines displayed this phenotype. The second of these phenotypes is an almost infallible diagnostic test for inaEN125. TrpP365 is a semidominant mutation in the trp gene, which encodes subunits of the TRP channel, the major carrier of the light-evoked current in Drosophila photoreceptors (45). The TrpP365 mutation makes the TRP channel constitutively open and kills the photoreceptor cells by allowing excessive Ca2+ entry. In TrpP365 homozygotes, the ERG amplitude is nearly zero even at 1 day post eclosion because of extensive photoreceptor degeneration. In TrpP365 heterozygotes (TrpP365/+), however, the time course of degeneration is much slower and a substantial ERG (10–15 mV) is present at 1 day post eclosion (table 9.3, row 1) and disappears with further ageing (hence, the mutation is said to be semidominant). If inaE is combined with TrpP365 (i.e., inaE/

7315.indb 208

11/15/07 1:42:16 PM

209


Table 9.3 Complementation Test between inaEN125 and 14959 P Insertion Mutants Genotype +/+;;P365/+ N125/+;;P365/+ 14959/+;;P365/+ N125/N125;;P365/+ N125/14959;;P365/+ 14959/14959;;P365/+

ERG

Amplitudes (mV)

14.0 ± 2.3 12.1 ± 1.5 13.5 ± 1.2 0.4 ± 0.3 3.0 ± 0.3 4.8 ± 0.7

(n = 9) (n = 6) (n = 9) (n = 10) (n = 6) (n = 6)

Note: All are white-eyed flies marked with either white (w) or cinnabar brown (cn bw) mutations. All ERGs were performed on 1 d.o. flies. TrpP365 and inaEN125 are abbreviated as P365 and N125, respectively.

inaE;;P365/+ for female), it enhances the aforementioned phenotype of TrpP365/+ so that only a very small ERG is present at 1 day post eclosion (table 9.3, row 4). We found that one of the P insertion mutants of the CG33147 gene, 14959, also had this property (table 9.3, row 6). Moreover, 14959 in heterozygous combination with inaE also displayed the same property; that is, inaE/inaE;;P365/+, inaE/14959;;P365/+ and 14959/14959;;P365/+ all had only a very small ERG at 1 day post eclosion (table 9.3, rows 4, 5, and 6) (41). Thus, 14959 P insertion allele and inaE are mutations in the same gene, or inaE and CG33174 are the same gene. The most definitive proof that a correct gene has been cloned is generally considered to be the restoration of the wild-type phenotype in the mutant in vivo through the introduction of the cloned gene sequence by germ-line transformation. We took advantage of the ability of the inaE mutation to enhance the Trp/+ phenotype, described earlier, to assess whether or not the introduced sequence rescues the mutant phenotype. Thus, TrpP365 heterozygotes (TrpP365/+) have 10–15 mV ERG responses at 1 day post eclosion. In the double mutant of inaE and TrpP365/+ (inaE/ inaE;;TrpP365/+), however, the responses are nearly zero at 1 day post eclosion because of the enhancement of the TrpP365/+ phenotype by inaE. When the rescue construct, consisting of the wild-type inaE cDNA sequence driven by a photoreceptor specific promoter, was introduced into the inaE/inaE;;TrpP365/+ double mutant, the transformant behaved like the TrpP365 heterozygote. That is, the rescue construct nullified the effects of inaE mutation in the double mutant, or rescued the inaE phenotype. Thus, all three lines of validation experiments unequivocally demonstrated that inaE and CG33174 are the same gene. 9.6.2.2 P222 and P255 We have identified two other new genes by the microarray approach, P222 and P255. Mutants in both these genes are impaired in synaptic transmission. Validation of identification is not as far advanced for these genes as for inaE. Nevertheless, the

7315.indb 209

11/15/07 1:42:16 PM

210


data obtained to date are sufficiently strong to lead one to believe that the identification is secure. There are four mutants falling into the P255 complementation group. Microarray experiments were carried out on two of these, JK1285 and JK1503. Deficiency mapping showed that P255 is uncovered by (included within) both the deficiencies Df(2L)VA6 and Df(2L)DS6, which have reported breakpoints at (37D2;38C6-E3) and (38E2;39E7), respectively. Thus, this mapping localized the P255 gene to the region of overlap between these two deficiencies, 38E2-3. However, the breakpoints of these deficiencies had not been determined at the molecular level. Therefore, we needed to allow for possible mismatches between the reported breakpoints of deficiencies and the locations of annotated genes. Again, we searched a much wider region than that determined by mapping—from 38D1 to 38F1, an approximately 220 kb region with 40 plus genes. When we examined the statistically analyzed microarray data for both JK1285 and JK1503 in this region, we found that the data from the two mutants were very different in quality. The data from the JK1503 mutant were noisy, with many genes of similar log-fold changes in expression and similar P values. Although these results suggested possible problems somewhere in the experimental steps, we made no attempts to repeat these data because the data from the other mutant were so clear-cut. In the data from the other mutant, JK1285, only two genes stood out in this entire region: CG9317 at 38E4 with a log-fold change of −1.31 and a P value of 0 and diaphanous (CG1768) at 38E7-8 with a log-fold change of −2.02 and a P value of 0. These were the only genes with log-fold changes >1.0 in absolute value and P values of 0 (table 9.4). The only other gene displaying a log-fold change >0.4 in absolute value was CG17470 at 38D2 (log-fold change: 0.65; P value: 1.98 × 10 −2). However, its P value was relatively large and its location at 38D2 was clearly outside the mapped limits of P255. To choose between CG9317 and diaphanous, we sequenced both cDNA and genomic sequences of both genes in JK1285 and JK1505 mutants as well as a wildtype control. There were no mutations in either the coding region or exon–intron boundaries of the diaphanous gene of either mutant, whereas there were mutations in the coding region of the CG9317 gene of both JK1285 and JK1503. JK1503 harbored a mutation that would introduce a stop codon near the 5′ end of the coding sequence. JK1285, on the other hand, harbored a mutation that is predicted to cause a Gly-Asp amino acid change within one of the transmembrane segments of the encoded protein. This mutation is likely to drastically alter the hydrophobicity of the transmembrane segment. These results strongly suggested that CG9317 is the P255 gene. As for P222, there are three mutants falling into the P222 complementation group: P222, P223, and JK1669. Microarray experiments were carried out on all three alleles and the corresponding wild-type controls. Earlier deficiency mapping had placed the P222 gene at 25D2-4;25D6-7 of the left arm of the second chromosome. This region happened to be well covered by deficiencies whose breakpoints have been determined at the molecular level. Remapping with some of these deficiencies placed the P222 gene between the right breakpoint of the deficiency Df(2L) Exel6011 and the left breakpoint of the deficiency Df(2L)cl7 (figure 9.2). The right breakpoint of Df(2L)Exel6011 at 25D5 has a molecular coordinate of 2L:5320589 (Flybase). The left breakpoint of Df(2L)cl7 at 25D7 is defined at the molecular level

7315.indb 210

11/15/07 1:42:16 PM

7315.indb 211

0.162140079 0.651517765 0.274181393 –0.12977622 –0.257087611 –0.141575284 0.304217116 0.136468743 –0.112077832 0.18375385 0.144351724 0.150378928 –1.3142205 0.128201465 0.210830067 0.166335291 –2.0230015 0.07940608 0.290662014 0.242106415 0.09671275 0.351549306 0.364121531

1627780_at 1641727_at 1635020_s_at 1640158_at 1639142_s_at 1639951_at 1629579_at 1640130_at 1639807_s_at 1627392_at 1630152_at 1625635_at 1634836_a_at 1636977_at 1635766_at 1635645_at 1625942_s_at 1627598_at 1636545_at 1623649_at 1635390_s_at 1623291_at 1638368_at

P Value 0.003849598 0.019801799 0.00014085 0.021034361 0.00104006 0.004522731 5.94E-09 0.018082278 0.001094611 0.015396777 0.008338214 0.001755247 0 0.008035779 0.003112032 4.68E-05 0 0.021928796 7.37E-06 0.000633153 0.001372121 4.52E-10 2.05E-06

Note: “Alignments” are as explained in table 9.1.

Log-Fold Change

Probe Set ID 0.24638223 1.244615801 0.307431775 0.250476964 0.340491415 0.219294899 0.203277575 0.256781459 0.149160382 0.336901807 0.241865569 0.209691477 0.38681972 0.213727089 0.312635319 0.172689394 0.29921406 0.154364841 0.269521319 0.306428811 0.13156216 0.213810033 0.315055654

SD

Table 9.4 Genes Showing Expression Changes in JK1285 (38D1-F1) Alignments 2L:20592770-20602014 2L:20612044-20612986 2L:20613089-20615531 2L:20615565-20617123 2L:20617274-20618592 2L:20619935-20620642 2L:20620496-20625918 2L:20628871-20645683 2L:20648662-20651586 2L:20678158-20678516 2L:2068145-2069086 2L:20693946-20696293 2L:20696884-20700051 2L:20700386-20704052 2L:20704656-20709521 2L:20720675-20722082 2L:20727888-20737265 2L:207398-210506 (+) 2L:20789892-20790609 2L:20791470-20793622 2L:20794177-20797486 2L:20799411-20800805 2L:20801219-20802525 — CG17470 phr6-4 CG2608 pncr 012:12L CG2611 CG2478 CG31678 CG1962 — — CG9316 CG9317 CG9318 Fs(2)Ket CG9319 dia CG11490 CheB38c — CG9331 CG31673 CG31674

Gene Symbol

38F1 38F1 38F1

38F1

E6 38E7-8

E4 E4 38E4-5

38D2 38D2 38D2 38D2 38D2 38D2 38D2-4 38D5

Cytol Position

Phototransduction in Drosophila 211

11/15/07 1:42:17 PM

212


25C8

25D5 P222

26A7 Df(2L)Exel6011 Df(2L)cl 7

Right break point of Df (2L)Exel6011 Left break point of Df (2L)cl 7 Molecular coordinate: 5320589 5346237 - 5365039

Figure 9.2 Deficiency mapping of P222. P222 was mapped between the right breakpoint of Df(2L)Exel6011 and the left breakpoint of Df(2L)cl7. The right breakpoint of Df(2L) Exel6011 is at 25D5 cytologically, and its molecular coordinate is 2L:5320589. The left breakpoint of Df(2L)cl7 is at 25D7 and is defined molecularly by the deficiency’s ability to disrupt or delete the nompC gene, located at 2L:5346237 to 5365039. Both the foregoing deficiencies complement P222. Another deficiency, Df(2L)Exel6012, with breakpoints at 25D5 and 25E6, uncovers the mutation; i.e., the mutation is within the deficiency. Deficiencies are shown as breaks in solid lines.

by the fact that the deficiency disrupts or deletes the nompC gene, which occupies the nucleotide coordinates 2L:5346237-5365039 (Flybase) (figure 9.2). The microarray data were somewhat noisier than those obtained with other mutants. Nevertheless, the data were clear in identifying the candidate gene in the mapped region. Table 9.5 displays all probe sets in the 25D4 to 25D7 region found to detect expression changes that are significant at a 0.05 level (see section 9.5) in the three mutants of the P222 complementation group. The probe set consistently detecting the largest log-fold change accompanied by a very low P value in each of the three mutants was that corresponding to CG14021. In addition, another probe set corresponding to cype showed moderately high log-fold changes and P values of 0 in the three mutants. A third one corresponding to vri had a relatively high log-fold change with a P value of 0 but only in one mutant, P223. Results of mapping, because they were defined at the molecular level, helped eliminate several of the probe sets shown in table 9.5. The last entry corresponding to the nompC gene was eliminated because nompC is disrupted by the Df(2L)cl7 deficiency (figure 9.2). The first four entries, from vri to CG14023, were eliminated because they lie at or to the left of the right breakpoint of the Df(2L)Excel6011 deficiency (figure 9.2). The right breakpoint of Df(2L)Excel6011 at 2L:5320589 runs through the CG14023 gene, located at 2L:5316240-5324927, and the other three genes, vri to CG14024 lie to the left of this breakpoint. Thus, the only two genes in the mapped region showing changes in expression of any significance in all three mutants were cype and CG14021 (table 9.5). We sequenced both these genes in all three mutants. There were no sequence alterations in the cype gene in any of the three mutants. By contrast, stop codons were found

7315.indb 212

11/15/07 1:42:18 PM

7315.indb 213

Log-Fold –0.680404453 –0.1920818 –0.279272584 –0.36860693 –0.45925195 –0.363730714 –0.539340969 –0.895430605 –0.447498813 –0.209368132

P Value 0 0.004021573 3.72E-05 1.99E-06 0 7.77E-15 3.38E-13 0 2.22E-16 0.000649058

P223

0.233696

0.261083184 0.223023798 0.220514924 0.429930948 0.208014883 0.252175926 0.313795286 –0.55860993

Log-Fold

P Value

0.0008696

1.53E-06 0.000136228 0.003405732 4.32E-07 0 3.98E-11 0.00136493 3.77E-15

JK1669

25D5 25D5-6 25D6 25D6 25D6 25D6 25D6-7 25D6-7

— CG14024 CG14023 cype CG14022 TotM CG14021 CG12512 nompC

Cytol Position 25D4-5

Gene Symbol vri

Note: Blank spaces in P222 columns correspond to probe sets that did not detect expression changes at the 0.05 significance level.

0 4.18E-07 5.41E-06 0 2.40E-05

–0.351956818 –0.197400154 –0.336016225 –0.593472788 –0.230068711

P Value

2.62E-13

Log-Fold

–0.281933617

Probe Set ID

1639273_s_at 1623674_at 1636447_at 1640735_at 1638416_at 1625430_at 1623635_at 1640365_s_at 1638636_at 1626936_at

P222

Table 9.5 Genes Showing Changes in Expression in P222, P223, and JK1669 (25D4-7)

Phototransduction in Drosophila 213

11/15/07 1:42:18 PM

214


in the coding region of CG14021 in all three mutants. These results strongly suggest that CG14021 is the P222 gene.

9.7 Concluding remarks In this chapter, we have reviewed a novel technique involving the use of DNA microarrays for cloning genes that are identified only by chemically induced ERGdefect-causing mutations. We have presented five applications of this approach, two to previously identified genes and three to new ones that had not been isolated previously. In all cases, the approach rapidly and reliably identified one or two candidate genes within the mapped interval, even though the quality of the microarray data varied with mutants. The main advantage of this approach is the speed with which the identification of a very small number of candidate genes can be achieved. It took, on the average, 3½ to 4 months from the time of fly head collection for RNA isolation to the time of identification of candidate genes. More than 50% of this time was taken up by the expansion of fly stocks, collection of heads, and RNA isolation. Had we used the two-cycle protocol for cRNA amplification (see section 9.4), which allows the use of as little as 10 ng amount of total RNA for microarray target preparation, the time of 3½ to 4 month could have been slashed by another month or so. To our knowledge, no currently available technique can match this feat. The speed is achieved by allowing one to skip the labor-intensive intermediate steps of fine mapping and proceed directly from deficiency mapping to the identification of a very small number of candidate genes. The final steps of gene identification and validation, on the other hand, are similar to those of other strategies. From the small number of candidate genes identified, the final identification of the single most likely candidate gene must be made, and the identification needs to be validated. These are achieved, as in other gene cloning strategies, by various combinations of gene sequencing, complementation with existing mutations, if available, and germ-line transformation. However, the present strategy does have one other advantage over other methods in that it provides information as to which of the candidate genes is likely to be the correct gene, allowing one to concentrate on it. By contrast, in other methods, all genes falling within the mapped interval must be considered equally. Microarray studies are not inexpensive. If we were to examine two allelic mutants in a gene, we need to prepare nine independent RNA samples, three each for the two mutants and a wild-type control. The cost of the nine chips required for these samples is $300 × 9 = $2700, and the charges of the Genomic Facility are $250 × 9 = $2250. In addition, an allowance needs to be made for the statistician’s time. However, these costs must be weighed against the time saved. If the target mutation happens to fall in a region rich in molecularly defined deficiencies and can be mapped within a region containing less than 5–6 genes by deficiency mapping alone, one could attempt to identify the gene directly by either sequencing or some other methods of mutation detection. However, it is still rare to achieve that kind of accuracy and resolution with deficiency mapping alone. One would need to map the mutant more accurately using the procedures we have outlined earlier (section 9.2.2). These procedures unfortunately can go on for months or even years. Our experience

7315.indb 214

11/15/07 1:42:18 PM


215

with the inaE gene is instructive in this regard. We had been attempting to clone this gene using the traditional methods for about 2 years with limited success. Had we applied the microarray strategy to this gene earlier, we estimate that we would have saved approximately 2 years of graduate student or postdoctoral time for this gene alone. The microarray costs, as exorbitant as they may seem, represent only a fraction of the money we spent on salaries. An even more important consideration than money is the invaluable time saved by using this approach.

Acknowledgments We thank Young Seok Hong for suggesting the use of the microarray approach for gene cloning and Kim Gilbert for help in the preparation of the manuscript. Some of the mutants mentioned in the paper were generated in other laboratories and generously given to us many years ago. The mutant inaEN125 was generated by Martin Heisenberg, and JK1285, JK1503, and JK1669 were generated by Jane Koenig and John Merriam. The first microarray experiment was done at the Center for Medical Genomics, Indiana University School of Medicine, Indianapolis, and all subsequent experiments were done at the Purdue Genomics Core Facility. We thank Phillip San Miguel, Ann Feil, and Fred Rakhshan of PGCF. This work was supported by grants from the National Eye Institute (EY00033) and National Institute of Mental Health (MH075041) to WLP.

Note

1. These investigators used the EP element for the generation of P insertion lines. The EP element is a P element containing a Gal4-responsive enhancer at one end (Rǿrth, 1996). If flies carrying an EP element are crossed to flies carrying Gal4 fused to a suitable promoter (46), in the progeny carrying both elements, an endogenous gene immediately adjacent to the EP element is induced. Thus those lines with EP insertions immediately upstream of a gene can be used for overexpression of the gene, and those with EP insertions in the antisense strand immediately downstream of a gene can be used for antisense knockdown of a gene.

References

7315.indb 215

1. Pak, W.L., Mutations affecting the vision of Drosophila melanogaster. In Handbook of genetics 3rd ed. King, R.C., ed., New York: Plenum, 703–733, 1975. 2. Pak, W.L., Study of photoreceptor function using Drosophila mutant. In Neurogenetics: Genetic approaches to the nervous system, ed. Breakefield, X., Amsterdam: Elsevier North, 67–99, 1979. 3. Pak, W.L., Drosophila in Vision Research: The Friedenwald Lecture, Invest Ophthalmol Vis Sci, 36: 2340–2357, 1995. 4. Pak, W.L. and Leung, H.T., Genetic approaches to visual transduction in Drosophila melanogaster. Receptor Channels, 9: 149–167, 2003. 5. Pak, W.L., Grossfield, J., and White, N.V., Nonphototactic mutants in a study of vision of Drosophila. Nature, 222(191): 351–354, 1969. 6. Hotta, Y. and Benzer, S., Abnormal electroretinograms in visual mutants of Drosophila. Nature, 222, 354–356 (26 April 1969); doi: 10.1038/222354a0,1969.

11/15/07 1:42:19 PM

216

7315.indb 216


7. Hotta, Y. and Benzer, S., Genetic dissection of the Drosophila nervous system by means of mosaics. Proc Natl Acad Sci U.S.A., 67(3): 1156–1163, 1970. 8. Heisenberg, M., Isolation of mutants lacking the optomotor response, Dros Inf Svc, 46: 68, 1971. 9. Heisenberg, M., Behavioral diagnostics: A way to analyze visual mutants of Drosphila. In Information processing in the visual systems of arthropods, ed. Wehner, R., Berlin: Springer-Verlag, 265–268, 1972. 10. Koenig, J. and Merriam, J., Autosomal ERG mutants, Dros Inf Serv 52: 50–51, 1977. 11. Zhai, R.G. et al., Mapping Drosophila mutations with molecularly defined P element insertions. Proc Natl Acad Sci U.S.A., 16;100(19): 10860–10865, 2003. 12. Parks, A.L. et al., Systematic generation of high-resolution deletion coverage of the Drosophila melanogaster genome. Nat Genet, 36, 288–292, 2004. 13. Ryder, E. et al., The DrosDel Collection: a set of P-element in Drosophila melanogaster. Genetics 167: 797–813, 2004. 14. Venken, K.J. and Bellen, H.J., Emerging technologies for gene manipulation in Drosophila melanogaster. Nat Rev Genet, 6(4): 340, Apr 2005. 15. Preston, C.R., Sved, J.A., and Engels, W.R., Flanking duplications and deletions associated with P-induced male recombination in Drosophila. Genetics, 144(4): 1623–1638, 1996. 16. Chen, B. et al., Mapping of Drosophila mutations using site-specific male recombination, Genetics, 149: 157–163, 1998. 17. Berger, J. et al., Genetic mapping with SNP markers in Drosophila, Nat Genet, 29: 475–481, 2001. 18. Schena, M. et al., Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science, 270: 467–470, October 1995. 19. Lockhart, D.J. et al., Expression monitoring by hybridization to high-density oligonucleotide arrays. Nat Biotechnol, 14(13): 1675–1680, December 1996. 20. Adams, M.D. et al., The genome sequence of Drosophila melanogaster, Science, 287: 2185–2195, 2000. 21. Bloomquist, B.T. et al., Isolation of a putative phospholipase C gene of Drosophila, norpA, and its role in phototransduction. Cell, 54(5): 723–733, August 26, 1988. 22. Montell, C. and Rubin, G.M. Molecular characterization of the Drosophila trp locus: A putative integral membrane protein required for phototransduction, Neuron 2(2): 1313–1323, April 1989. 23. Schneuwly, S. et al., Drosophila ninaA gene encodes an eye-specific cyclophilin (cyclosporine A binding protein), Proc Natl Acad Sci U.S.A., 86(14): 5390–5394, July 1989. 24. Li, C. et al., INAF, a protein required for transient receptor potential Ca2+ channel function. Proc Natl Acad Sci U.S.A., 96: 13474–13479, 1999. 25. Burg, M.G. et al., Genetic and molecular identification of a Drosophila histidine decarboxylase gene required in photoreceptor transmitter synthesis, EMBO J, 12: 911–919, 1993. 26. Burg, M.G. et al., Drosophila rosA gene, which causes aberrant photoreceptor oscillation, encodes a novel neurotransmitter transporter homologue, J Neurogenet, 11: 59–79, 1996. 27. Geng, C. et al, Target of Drosophila photoreceptor synaptic transmission is histaminegated chloride channel encoded by ort (hclA). J Biol Chem, 277, 42113–42120, 2002. 28. Lee, H.S. Personal communication. 2004. 29. Kornberg, T.B. and Krasnow, M.A., The Drosophila genome sequence: Implications for biology and medicine. Science, 287(5461): 2218–2220, 2000. 30. Ishikawa, H., Evolution of ribosomal DNA, Comp Biochem Physiol, 58(1): 1–7, 1977. 31. Craig, B.A., Black, M.A., and Doerge, R.W., Gene expression data: The technology and statistical analysis. J Agric Biol Environ Statistics (JABES) 8(1): 1–28, 2003.

11/15/07 1:42:19 PM


217

32. Benjamini, Y. and Hochberg, Y., Controlling the false discovery rate: a practical and powerful approach to multiple testing, J R Stat Soc, Series B, 57: 289–300, 1995. 33. Holm, S., A simple sequentially rejective multiple test procedure, Scand J Statistics, 6: 65–70, 1979. 34. Cosens, D.J. and Manning, A., Abnormal electroretinogram from a Drosophila mutant, Nature, 285–287; doi: 10.1038/224285a0, Oct 18, 1969. 35. Minke, B., Wu, C., and Pak, W.L., Induction of photoreceptor voltage noise in the dark in Drosophila mutant. Nature, 258(5530): 84–87, 1975. 36. Hardie, R.C. and Minke, B., The trp gene is essential for a light-activated Ca2+ channel in Drosophila photoreceptors. Neuron, 8(4): 643–651, 1992. 37. Wong, F. et al., Proper function of the Drosophila trp gene product during pupal development is important for normal visual transduction in the adult, Neuron, 3(1): 81–94, 1989. 38. Montell, C., The TRP superfamily of cation channels. Sci STKE, 2005(272): re3. Review, February 22 2005. 39. Minke, B., TRP channels and Ca2+ signaling. Cell Calcium, 40(3): 261–275. Review, September 2006. 40. Hardie, R.C., TRP channels and lipids: From Drosophila to mammalian physiology. J Physiol. 578: 9–24, 2007. 41. Leung, H-T et al., unpublished results, n.d. 42. Postlethwait, J.H. and Handler, A.M., The roles of juvenile hormone and ecdysone during vitellogenesis in isolated abdomens of Drosophila melanogaster. J Insect Physiol, 25: 455–460, 1979. 43. Barnett, T.C. et al., The isolation and characterization of Drosophila yolk protein genes. Cell, 21: 729–738, 1980. 44. Hovemann, B. et al., In Drosophila melanogaster: Sequence of the yolk protein I gene and its flanking regions. Nucl Acids Res, 9(18): 4721–4734, September 25 1981. 45. Yoon, J. et al., Novel mechanism of massive photoreceptor degeneration caused by mutations in the trp gene of Drosophila. J Neurosci, 20(2): 649–659, 2000. 46. Brant, A.H. and Perrimon, N., Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development 118(2): 401–415, 1993.

7315.indb 217

11/15/07 1:42:20 PM

7315.indb 218

11/15/07 1:42:20 PM

4 Insulin Receptor-Based Signaling in the Vertebrate Retina

7315.indb 219

11/15/07 1:42:31 PM

7315.indb 220

11/15/07 1:42:31 PM

10

Insulin Receptor-Based Signaling in the Retina Patrice E. Fort, Ravi S.J. Singh, Mandy K. Losiewicz, and Thomas W. Gardner

Contents 10.1 10.2

Introduction................................................................................................. 221 Expression and Phosphorylation of the Insulin Receptor........................... 223 10.2.1 Receptor Autophosphorylation..................................................... 223 10.2.2 Receptor Ligand Affinity..............................................................224 10.3 Functional Assays.......................................................................................224 10.3.1 IR Kinase Assay...........................................................................224 10.3.2 Kinase Assays for Other Proteins of the IR Signaling Pathway... 225 10.3.2.1 Akt-1 Kinase Assay...................................................... 225 10.3.2.2 PI3-Kinase Assay......................................................... 227 10.3.3 Potential Pitfalls............................................................................ 228 10.3.4 Bioluminescent Assay................................................................... 229 10.4 Retinal Specificity Compared to Other Insulin-Sensitive Tissues............. 230 10.4.1 In Situ Hybridization.................................................................... 230 10.4.1.1 Akt-1, -2, and -3 in Situ Hybridization......................... 231 10.4.2 Laser Capture Microscopy............................................................ 231 10.5 General Approaches.................................................................................... 232 10.5.1 Proteomic Analysis....................................................................... 232 10.5.2 Kinome Analysis.......................................................................... 233 10.6 Conclusions................................................................................................. 233 Acknowledgments................................................................................................... 234 References............................................................................................................... 234

10.1 Introduction The retina is one of the numerous insulin-sensitive tissues. Studies conducted in mice, rats, and rabbits have shown that the insulin receptor (IR) has a high basal kinase activity in retina that does not fluctuate with feeding or fasting. By contrast, IR activity in the liver varies markedly with fasting (low plasma insulin levels, 221

7315.indb 221

11/15/07 1:43:00 PM

222


low IR activity) and feeding (high plasma insulin levels, high IR activity). The IR, along with insulin-like growth factor-1 receptor (IGF-1R) and insulin receptor-related receptor (IRR), belongs to a family of heterotetrameric (α2β2) transmembrane glycoprotein receptors that are widely expressed in mammalian tissues. Although little is known about the IRR, the IR and IGF-1R are the products of different genes that lead to the expression of proreceptors that display more than 50% amino acid sequence identity. The IR family ligands include three structurally related peptides that also present a high amino acid sequence identity: insulin, IGF-1, and IGF-2. Posttranslational processing results in the dimerization and disulfide linkage of proreceptors followed by proteolytic cleavage that generates α and β subunits (1). The extracellular α subunit contains the ligand-specific binding site, and the transmembrane β subunit possesses the tyrosine kinase activity. The kinase activity is induced by the interaction of the ligand with the α subunit that induces the transautophosphorylation of the β subunits. Despite their high degree of identity, insulin and IGFs can trigger very different functions. Insulin is mostly involved in the regulation of the metabolic functions of the classically insulin-responsive tissues, liver, adipose, and skeletal muscle. Evidence also suggests that insulin acts on neural tissue and can modulate neural metabolism, synapse activity, and feeding behaviors (2). Insulin receptors are expressed on both the vasculature and neurons of the retina, but their functions are not completely defined. IGFs seem to play essential roles in the regulation of growth and development in different tissues as shown by different transgenic models. However, in the central nervous system, including the retina, evidence suggests that insulin action stimulates neuronal development, differentiation, growth, and survival, rather than acutely stimulating nutrient metabolism; e.g., glucose uptake as in skeletal muscle (3–6). IR from retinal neurons and blood vessels share similar properties with IR from other peripheral tissues, and retinal neurons express numerous proteins that are attributed to the insulin signaling cascade, as in other tissues. However, undefined neuron-specific signals downstream of the insulin receptor are also likely to exist. The retina and brain are unique in containing a high number of different cell types and being heterogeneous compared to other insulin-sensitive tissues such as liver and skeletal muscles. Although the distinct physiological functions of insulin and IGFs depend on differences in the distribution and/or signaling potential of their respective receptors, the various retinal cell types can differentially express these receptors and thus have varying sensitivity to these ligands. Moreover, IR and IGF-1R receptors are able to assemble not only as homodimers but also as heterodimeric hybrid structures containing an (αβ) half of the IR disulfide-linked to an (αβ) half of the IGF-1R in different tissues and cell lines (7, 8). This feature adds another level of complexity to the function of these pathways and their sensitivity to the different ligands. In this chapter, we will detail how the expression, and more importantly, the activity of the different proteins of these pathways, can be assessed in the retinal tissue. We will particularly discuss the various parameters of these assays that can be modified. We will finally discuss two different methods that can be used to characterize more deeply the signaling cascades involved in these pathways.

7315.indb 222

11/15/07 1:43:01 PM

Insulin Receptor-Based Signaling in the Retina

223

10.2 Expression and Phosphorylation of the insulin receptor 10.2.1 Receptor Autophosphorylation The first step of the activation of both insulin and IGF pathways is the trans-autophosphorylation of the receptors induced by ligand binding. This step is necessary for the activation of the kinase activity of the β subunits of both receptors. Note that the IGF-2 receptor does not possess tyrosine kinase activity, and IGF-2 stimulates the IGF-1R and IR. Characterization of the phosphorylation state can be assessed using specific antibodies directed against either IRβ or IGF1-Rβ (Biotechnology, Santa Cruz, California) to immunoprecipitate the receptors and purify them from original retinal lysates (9). Lysates are prepared in IP buffer (50 mM HEPES, 137 mM NaCl, 1 mM MgCl2, 1 mM CaCl2, 2 mM EDTA, 2 mM NaVO4, 10 mM NaF, 10 mM sodium pyrophosphate, 10 mM benzamidine, 2 mM phenylmethylsulfonyl fluoride, 10% glycerol, 1% NP-40, and protease inhibitor tablet; Roche, Mannheim, Germany).

7315.indb 223

1. Preincubate 5 µL of antibody with 50 µL of 50% slurry protein A equilibrated in PBS containing 1% of BSA in 500 µL of IP buffer, for 4 h at 4°C to allow the interaction of the antibody with the beads. 2. Centrifuge for 1 min at 500 × g, and discard the supernatant. 3. Wash the beads pellet with 250 µL of IP buffer. (Repeat steps 2–3, twice.) 4. Add 500 µg of protein lysate in 1 mL of IP buffer final, and incubate overnight at 4°C rocking to allow the interaction of the prebound antibody with its epitope. 5. Centrifuge for 1 min at 500 × g, and discard the supernatant. 6. Wash the beads pellet with 500 µL of IP buffer. (Repeat steps 5–6, twice.) 7. Add 30 µL of 2× sample buffer (0.13 M Tris, pH 6.8, 25% glycerol, 2.5% β-mercaptoethanol, 2% SDS, 2.3 mg/mL bromophenol blue) and boil for 5 min before proceeding with SDS-polyacrylamide gel electrophoresis. 8. Centrifuge for 2 min at 2000 × g to pellet the beads, and run the supernatant on 4%–12% Bis–Tris gel in MOPS buffer at 200 V for 55 min. 9. Proteins are then electrotransferred to nitrocellulose membrane for 2 h at 400 mA at 4°C. 10. Block the membrane with TBST containing 3% of BSA for at least 1 h. 11. Incubate primary antibody antiphosphotyrosine (Upstate, Charlottesville, Virginia) in blocking solution overnight at 4°C. 12. Wash thrice for 5 min each in TBST. 13. Incubate horseradish peroxidase-conjugated secondary antibody 1 h at room temperature in the blocking solution. 14. Wash thrice for 5 min each in TBST. 15. Add detection reagents (ECL+, Amersham, Piscataway, New Jersey) onto the membrane, and incubate for 5 min at room temperature. 16. Expose autoradiography film on top of the membrane and develop.

11/15/07 1:43:01 PM

224


10.2.2 Receptor Ligand Affinity Insulin stimulates the type 1 IGF receptor homodimer as well as the IR–IGF-1R heterodimer in different cell types, including retinal cells during development, although it has a much lower affinity for the IGF-1R than for IR (10). This has been shown using binding assays that measure the displacement of radiolabeled ligand by addition of unlabeled ligand. This technique can be used on retinal explants or retinal cells in culture.

1. Cells are incubated for 4 h at 10°C in HEPES binding buffer, pH 7.8 (100 mM HEPES, 120 mM NaCl, 5 mM KCl, 1.2 mM MgSO4, 8 mM glucose, and 0.1% bovine serum albumin) with the addition of 12.5 × 10 −12 M of 125I-insulin or 125I-IGF-1 and increasing concentrations of unlabeled insulin or IGF-1. 2. The cells are then washed 4 times with ice-cold PBS. 3. The cells are then homogenized in 0.1% SDS, and the radioactivity is measured in a gamma counter. 4. Unspecific binding is defined as the binding of radiolabeled ligand in the presence of excess unlabeled ligand.

This method allowed Nitert et al. (8) to show that human endothelial cells have a larger amount of IGF-1R than IR at their surface because the specific binding of IGF-1 is more than twofold higher compared to insulin. The same technique is used on coimmunoprecipitated dimers formed by IR and IGFR to characterize their respective ligand affinity. As mentioned earlier, IR and IGFR can form heterodimers in several tissues and cell types in addition to homodimers, which have different properties that add another level of regulation of the IR and IGF-1R pathways.

10.3 Functional assays The insulin and insulin-like growth factor pathways are composed of a series of phosphorylation-activated kinases following the receptor–ligand interaction.

10.3.1 IR Kinase Assay The first step in the insulin signaling pathway is tyrosine autophosphorylation of the receptor inducing the activation of the kinase domain of the IR. Insulin receptor kinase activity depends on autophosphorylation, as well as serine phosphorylation and adaptor proteins, such as Grb14 and Src. The β subunit of the IR is immunoprecipitated using the same antibody mentioned in section 10.2.2.

7315.indb 224

1. Preincubate 5 µL of antibody with 50 µL of slurry protein A equilibrated in PBS containing 1% of BSA in 500 µL of IP buffer for 4 h at 4°C to allow the interaction of the antibody with the beads. 2. Centrifuge for 1 min at 500 × g and discard the supernatant. 3. Wash the beads pellet with 250 µL of IP buffer. (Repeat steps 2–3, twice.)

11/15/07 1:43:01 PM


225

4. Add 500 µg of protein lysate in a final volume of 1 mL of IP buffer and incubate overnight at 4°C rocking to allow the interaction of the antibody with its epitope. 5. Centrifuge at 500 × g for 1 min, and discard the supernatant. 6. Add 500 µL of kinase buffer (50 mM HEPES, pH 7.3, 150 mM NaCl, 20 mM MgCl2, 2 mM MnCl2, 0.1% Triton X-100, and 0.05% bovine serum albumin). (Repeat steps 5–6, twice.) 7. Add 200 µL of kinase mix (112.5 µM ATP, 4.5 mg/mL polyGlu:Tyr, 0.25 µCi/ mL 32P–ATP in kinase buffer), and rock for 60 min at room temperature. 8. Pulse spin the samples to stop the reaction, and transfer 40 µL on filter paper. Let it dry for 15 s, and then place the filter paper in 0.75% phosphoric acid. 9. Wash filters 5 times for 5 min each in 0.75% phosphoric acid and once in acetone. Air-dry filters, and transfer in tubes with 5 mL of scintillation liquid.

Using this technique, our group was able to show that retinal IR has a high constitutive kinase activity, equivalent to that of postprandial liver, but it does not fluctuate with feeding or fasting (11). Comparison of kinase activities from control and diabetic rats retinas also enabled our group to show that diabetes induces a 25% decrease in the IR activity as early as 4 weeks after the onset of diabetes (12), whereas the tyrosine phosphorylation decrease was not yet detectable (figure 10.1a) before 12 weeks of streptozotocin-induced diabetes (figure 10.1b).

10.3.2 Kinase Assays for Other Proteins of the IR Signaling Pathway The activity of several other downstream kinases of the insulin signaling pathways can be tested, including Akt–1, Akt–3, and PI3–kinase, which are activated following IR activation. Here, we will describe two different kinase assays. 10.3.2.1 Akt-1 Kinase Assay

7315.indb 225

1. Preincubate 5 µL of antibody with 50 µL of slurry protein G in 50 µL of buffer A (2 M Tris, pH 8, 0.5 M EDTA, 1 M EGTA, 0.5 M sodium fluoride, 0.5 M sodium β-glycerophosphate, 0.5 M sodium pyrophosphate, 0.5 M sodium orthovanadate, 10% Triton X-100, 0.1% β-mercaptoethanol, 1 mM LR-microcystin, and proteinase inhibitors cocktail), standing for 1 h at 4°C to allow the interaction of the antibody with the beads. 2. Centrifuge for 1 min at 500 × g, and discard the supernatant. 3. Wash the beads pellet with 500 µL of buffer A. (Repeat steps 2–3, twice.) 4. Add 500 µg of protein lysate in a final volume of 500 µL of buffer A and incubate for 1 h at 4°C, rocking to allow the interaction of the antibody with its epitope. 5. Centrifuge at 500 × g for 1 min, and discard the supernatant. 6. Wash the beads pellet with 500 µL of buffer A (thrice), then twice with buffer B (2 M Tris, pH 7.5, 30% Brij 35, 1 M EGTA, 0.1% β-mercaptoethanol) and once with ADBI buffer (0.5 M MOPS, 0.5 M sodium β-glycerophosphate, 1M EGTA, 0.5 M sodium orthovanadate, 0.5 M DTT).

11/15/07 1:43:01 PM

226

Signal Transduction in the Retina 150 CTRL

DIAB

BLOT: PY BLOT: IRβ

Percent Control

IP: IRβ

CTRL DIAB

125 100 75 50 25 0

PY/IRβ

Total IRβ

IR Kinase Activity, % Control

(a) 120 100

*

80 60 40 20 0

CTRL

DIAB (b)

DIAB + INS

Figure 10.1 Reduced retinal insulin receptor phosphorylation and kinase activity in diabetes. The insulin receptor (IR) from retina was immunoprecipitated (IP) and analyzed for tyrosine phosphorylation (PY) or kinase activity. (a) Analysis of immunoblots for tyrosine phosphorylation of the insulin receptor-β (IRβ) and total insulin receptor-β reveals equivalent phosphorylation and expression in the retinal tissue of control (CTRL) and 4-week-diabetic (DIAB) rats (representative immunoblots shown, n = 8 control and 5 diabetic rats). (b) Four weeks of diabetes (n = 26) reduces insulin receptor kinase activity compared with controls (n = 27) (P < 0.001). Retinal insulin receptor kinase activity was normal in diabetic rats treated with insulin pellets (INS) (n = 14) when compared with controls (P = 0.12). Retinal insulin receptor kinase activity was not significantly reduced in 4-week-old diabetic rats when compared with diabetic rats treated with insulin pellets (P = 0.23); however, treatment with insulin pellets trended toward improved insulin receptor kinase activity.

7. Add 40 µL of kinase mix (40 μM PKA, 0.4 mM Akt/SGK Substrate Peptide, 112.5 µM ATP, 0.25 µCi/µL 32P-ATP) and rock for 60 min at 30°C. 8. Pulse spin the samples to stop the reaction, and transfer 40 µL on filter paper. Let it dry for 15 s, and then place the filter paper in 0.75% phosphoric acid. 9. Wash filters 5 times for 5 min each in 0.75% phosphoric acid and once in acetone. Air-dry filters, and transfer in tubes with 5 mL of scintillation liquid.

Using this method, we have shown that the Akt-1 and Akt-3 activities were reduced in rat retina as early as 4 weeks after the onset of diabetes, whereas the expression and phosphorylation level were decreased after 12 weeks of diabetes (figure 10.2) (12).

7315.indb 226

11/15/07 1:43:02 PM

227

Insulin Receptor-Based Signaling in the Retina CTRL BLOT: Akt?? IP:Akt-1 BLOT:

CTRL

DIAB

P-Akt??

BLOT: P-Akt?? BLOT:

Akt?? (a)

Akt-1 Kinase Activity % Control

BLOT:

DIAB

P-Akt??

120 100 80

*

60 40 20 0

CTRL

DIAB (b)

DIAB + INS

Figure 10.2 Reduced Akt kinase activity by diabetes in rat retina. (a) Retinal lysates from control (CTRL) and diabetic (DIAB) rats were analyzed by immunoblot analysis. In retina, Akt phosphorylation of threonine (Thr) 308 and serine (Ser) 473 and total expression were unaltered in the diabetic state. (b) After 4 weeks of diabetes, Akt-1 kinase activity in retina was reduced by 54% (*P < 0.01 by ANOVA and Tukey–Kramer multiple comparisons post hoc test, n = 7/group).

10.3.2.2 PI3-Kinase Assay Phosphatidylinositol 3-phosphate (PI3-K) is involved in numerous different mechanisms in cell activity. Hence, the assay for PI3-K activity specifically involved in the insulin signaling pathway is based on its interaction with the specific docking proteins called IRS-1 and IRS-2 for insulin receptor substrate.

7315.indb 227

1. Preincubate with 2 µg each of anti–IRS-1 and anti–IRS-2 (Santa Cruz Biotechnology, Santa Cruz, California) with 30 µL of 50% protein G Sepharose slurry (Amersham) in 500 µL of IP buffer for 4 h. 2. Centrifuge for 1 min at 500 × g, and discard the supernatant. 3. Wash the beads pellet with 500 µL of IP buffer. (Repeat steps 2–3, twice.) 4. Add 75 µg of protein lysate in 1 mL of IP buffer final, and incubate overnight at 4°C rocking. 5. The immune complexes are washed once with buffer A containing 0.5 mol/L NaCl, once with buffer B (50 mmol/L Tris-HCl, pH 7.5, 0.03% Brij-35 [vol/ vol], 0.1 mmol/L EGTA, and 0.1% β-mercaptoethanol [vol/vol]), and once with TNE buffer consisting of 20 mmol/L Tris-HCl, pH 7.5, 100 mmol/L NaCl, 0.5 mmol/L EGTA, and 0.1 mmol/L sodium orthovanadate. 6. The immune complexes are then incubated at 35°C for 10 min in 50 µL TNE buffer, pH 7.4, in the presence of 32P-ATP (10 µCi/assay) and the substrate phosphatidylinositol (20 µg/assay). 7. The reaction is stopped by adding 20 µL of 6 N HCl and 160 µL of CHCl3/ CH3OH (1:1). 8. The organic phase is spotted on a thin-layer chromatography plate and subjected to ascending chromatography using the solvent CHCl3/CH3OH/H2O/ NH4OH (60:47:11.3:2).

11/15/07 1:43:03 PM

228


9. Phosphatidylinositol 3-phosphate, thus resolved, is quantified by PhosphorImager analysis (Molecular Dynamics, Sunnyvale, California).

With this assay, our group was able to show that IRS2-dependent PI3-K activity was reduced in the retina after 4 weeks of diabetes, whereas the level of expression and phosphorylation were unchanged (figure 10.3) (12).

10.3.3 Potential Pitfalls The efficiency of IP-based kinase assays depends on multiple parameters that can be optimized according to the nature of the targeted kinase. This section will concentrate on these diverse parameters.

Figure 10.3 Reduced IRS-1/2–PI3-K activity in retina of diabetic rats. Retinal lysates were dual-immunoprecipitated for IRS-1/2, and PI3-K activity was measured. A representative TLC plate image is shown demonstrating reduced phosphatidylinositol 3-phosphate (PI3P) formation in lysates of 4-week-old diabetic rats (n = 14 control [CTRL] and 15 diabetic [DIAB] rats), with corresponding graphic representation.

7315.indb 228

11/15/07 1:43:03 PM


229

The first parameter to be optimized is the protein amount used for the IP, which will affect the intensity of the final signal. The amount of protein used to perform the IP must be adapted to the relative concentration of the kinase in the original lysate and the affinity of the antibody. Although IR and Akt-1 proteins are expressed at modest levels in the retina, we used a high amount of total retinal lysate to perform the IPs and found that 500 µg of protein leads to a good signal-to-noise ratio. This ratio also depends on the efficiency of the washes performed during the IP and after the substrate phosphorylation. The step of prebinding the antibody with the beads, and the washes following are especially important with weakly expressed proteins, because they reduce the loss of protein due to binding to free antibody. The series of washes performed after the lysate incubation is necessary to decrease the background activity induced by kinases that nonspecifically interact with the beads. During the optimization process, we determined that the number of washes was even more crucial than the volume of buffer used; three washes of 500 µL of IP buffer were more efficient than two washes of 1 mL. Finally, we also determined that at least five washes were required following the substrate incubation to decrease nonspecific binding of the free radioactive ATP to the filter paper. Another parameter to be optimized is the duration of substrate incubation, which determines the intensity of the final signal obtained. The duration of the reaction is correlated to the abundance and the activity of the kinase in the lysate tested. In the retina, we have determined that a 1 h reaction is the optimum duration to obtain the best signal-to-noise ratio for IR and Akt1 kinases. Key components of the immunoprecipitation-based kinase assays, especially those using nonspecific substrates, are the negative controls without antibody and, to a lesser extent, without lysate to prove the specificity of the approach. The so-called no-lysate control is necessary to show that the signal obtained for the IPs of interest is due to a specific interaction of the specific antibody used and not due to any nonspecific binding to the beads.

10.3.4 Bioluminescent Assay (13) Another way to test the activity of both IR and IGF-1R uses a technology called bioluminescence resonance energy transfer (BRET). Contrary to assays previously mentioned that allow measurement from tissue samples, this assay can only be used in cells and allows the measurement of the IR or IGF-1R activity in living cells in real time. Moreover, this assay does not measure the kinase activity but the conformational change following substrate activation of the receptors necessary for transautophosphorylation. Because transfecting primary retinal cells remains particularly challenging, this technique is of more interest for use in retinal cell lines such as R28 or RGC-5. The principle of the method is to transfect cells with two different constructs: one is fused to the 35 kDa Renilla luciferase (RLuc, the donor), and the other is fused to the 27 kDa enhanced yellow fluorescent protein (EYFP, the acceptor). The synthetic RLuc substrate, coelenterazine h, is able to permeate the cell membrane and is converted by RLuc under the emission of blue light of 480 nm. When EYFP is in close proximity (

Signal Transduction in the Retina

Signal Transduction

Signal Transduction